Too Much Faith Is Being Placed In Untested Theories

Submitted by Peter Tchir of TF Market Advisors,

A Pseudoscience Stuck in Place

I am growing more concerned by the day by the actions of the central banks.  It isn’t just that markets popped and dropped dramatically before and after Draghi’s rate cut, or that any policy seems particularly bad, just that the policies don’t seem to be working great, and are leaving a changed landscape that will need to be corrected, somehow, in the future.  I am quite simply concerned that too much faith is being placed in untested theories that may or may not work, or may or may not even be correct.

Here are a few things that concern me the most:

1. Central Bankers seem to rely on economic theories that have remained largely unchanged for years


2. Central Bankers seem of an age that they aren’t willing to incorporate theories that might change their favored models or might make those models too complex to be easily understood by those in charge (the Nobel prize committee has given out multiple awards for work in behavioral economics, yet central bank models seem to rely on pretty basic econometric models where behavior doesn’t rapidly change based on policies)


3. Central Bankers seem focused on domestic issues without really considering the ramifications and ripple effects that they potentially create

From Newton to Bohr

I liked Newtonian physics.  I could do the math easily and it was intuitive.  It was so easy that I took physics 101 right along with econ 101 because I needed some easy A’s.

But physics has changed.  The relatively simple world of Newtonian physics turned out to be inadequate to explain what was needed to propel science forward.  As comforting as it was to know that “each and every action has an equal and opposite reaction” that just doesn’t cut it in high end physics.

Personally, I started to lose interest and any real intuitive skills in physics around the time we learned that light is both a wave and a particle.  The math was getting more complex, but I could muddle my way through that.  What I lost was any instinctive or intuitive feel for what was being modeled.  I tended to focus on areas that I felt comfortable with, hampering any potential for intellectual growth.

Quantum mechanics really revolutionized physics. It was a new paradigm and you either had to adapt, understand it, or get some intuitive feel because brute force math might be enough to be adequate in the field, but not enough to excel.

I wonder why economics hasn’t had its “quantum mechanics” moment?

Did Keynes and Hayek really discover all there is to know?  Is Yellen’s beloved “Taylor rule” really the epitome of the “scientific” advancement of economics?  I realize there have been advances, but most seem to be “more of the same”.  No one seems to have challenged the central tenants of macroeconomics.

In physics, once Newtonian physics failed to explain the world, brilliant minds concocted experiments to test hypothesis.  This is what led to quantum mechanics.  The old theory was failing in that it couldn’t explain some observed phenomena, so it was ultimately supplanted by a new, much more complex theory, but one that explained much more of what was observable.

Why is that not happening in economics?  Personally, I don’t think economics has done a great job in explaining the world, otherwise we shouldn’t have so many periods of boom and bust globally.

Maybe it is the inability to experiment?  This is potentially a bigger issue than it seems at first.  We do experiment in economics, but it is a small group of elite, and mostly collegial economists who get to experiment.  Actually they get to put their beliefs into practice and then argue that the situation is better than if they hadn’t been allowed to implement their theories.  While costs and access can stop scientific progress, there certainly was a time that it was more readily available.  Hypothesis could be tested and failures catalogued and successes expanded on AND verified by repetition.  This capability just doesn’t exist in macroeconomics.  There are NO TWO economies that are identical except for the policies implemented.

Young professors could and did challenge the system in the hard sciences.  It is probably no co-incidence that most scientific Nobel prizes are awarded for work “conceived of” when the person was in their 20’s and “performed” in their 30’s.  That might be a generalization, but it isn’t entirely inaccurate.

Maybe economics is failing to attract good new people?  There may be something to this.  To some extent the economists that I know and respect the most (yes, I do like and respect some economists) had strong quantitative skills but an interest in business.  The didn’t want to be a “math” geek and liked working with the “real world”.  I am willing to make the conjecture that as computer science grew and the opportunities there grew, it was an even better match for many quantitative students who wanted something other than pure math, or physics, or chemistry, than economics.  Maybe even as MBA’s started looking for more “quants” even more people who would be the new economists didn’t pursue that?

Or maybe economists just ignore their own?  I have read a little about “behavioral economics”.  My take is that it demonstrates that people don’t always do what would produce the best “expected outcome”.  That the “rational man” that economic models are built on may not exist, and what is rational on a purely “economic” level might not be applicable on an overall evaluation.  We tend to hate losses more than we like winning.  How is that incorporated into the econometric global macro models used by the Fed?  The Fed runs the treasury/dividend yield model.  Yeehaw, except for the graphs that is nothing a good old fashioned HP12C couldn’t handle.  Why aren’t we incorporating some new techniques?  Maybe, because just like I hit the wall in physics at a certain point, the economists in charge have no interest in trying to incorporate things into their models that they don’t intuitively understand, might call into question their own body of “prestigious” work, and where quite frankly they might not have the technical expertise needed to incorporate them?

The Observer Effect.  Science understands that the act of observation can actually impact whatever is being observed.  Attempting to measure something affects the measurement.  First, I question how that plays into anything that is a “survey” or that is “subjective” in the first place.  How many purchasing managers hoping for better year-end bonuses say things are better than they are because they know their boss will like it, and at this point, they know the stock market will like it.  What about the “household survey” for non farm payrolls that we will get tomorrow.  Does it make a difference how you respond depending on your political party?  Does it amaze you that we still conduct door to door surveys to figure out how many Americans are working?  This is all a relatively minor effect, yet probabl
y real, and as far as I can tell, largely ignored at the “policy” level.

Learned Behavior.  Humans learn over time.  We are pretty adept and maximizing return while minimizing risk.  This is where I think economics does the worst job of integrating its own new theories.  QE seems based on a pretty simplistic model.  Provide more money, take risky investments out of the market, and the market will take that money and be encouraged to take more risk.  It will create asset price inflation which will encourage further real risk taking.  What if it turns out it is easier not to take the risk but end up with a pretty darn good reward?  How many companies took risk and a lot better off for it?  But how many have decided it is easier to do some financial engineering and let QE take care of their stock price?  How is that accounted in the models?  It probably isn’t and is probably too complicated, but we don’t try and predict the weather by licking our finger and sticking it in the air, yet economists seem in many ways content to run their policy on little more than that.

Equal and Opposite Reactions.  Such a basic concept.  It extends beyond science.  If you punch someone in the face, you can reasonably expect a certain reaction.  You might be able to qualify even that reaction based on the size and personality of the person you punch in the face.  Then why don’t we seem to use that in economics?  We live in a global economy apparently some of the time, but inflation is local wage driven only?  Hmm.  Bernanke, who claims protectionism was part of the problem with the Great Depression, basically told the Emerging Market countries (through lackey’s in Jackson Hole) that we will do what we need and they can worry about themselves.  Draghi cut rates today.  What does that to their currency?  Does that help or hurt what we have been trying to work on?  Central bankers all too often seem to act as though they are playing golf when the game is really chess.

Kasparov to Big Blue

Which brings us to chess.

Maybe the central bankers are aware that they are playing chess.  Maybe Bernanke is aware that each of his moves will cause another move by his “opponent” which he will then have to react to.  The problem is that if he is playing chess, and he is “thinking a few moves ahead” he is assuming too much, and making a classic mistake of expecting his opponent to fall into his “traps”

In the early days of “computer” chess, a modestly better than average human player could beat most computers after a few matches.  That was because of how computers evaluated the chess board.  There were far too many moves for a computer to analyze all the possibilities.  So they used “heuristics” to “score” boards.  They found ways to estimate how strong or weak a position was.  They could then “truncate” paths that lead to weak positions and explore only “strong” paths. The trick was figuring out what the computer was doing incorrectly.  To take it down a path that looked “strong” for several moves that could be then turned around.  The computer literally “fell for the trap”.

But “Big Blue” changed that.  It literally was so fast that it didn’t have to “truncate” paths early.  It could play out 200 million positions in a second and ultimately beat Kasparov.  That was back in 1997.

It was a sad day for many since it turned something that was elegant with a certain flair where imagination was respected and turned it into a brute force mechanical process.

I am not arguing that economics is something that is purely brute force, but I do think there are two lessons to be drawn from this:

1. Computer power and the evolution and development of computer power to analyze complex systems is useful and I am not sure we do enough of that, and


2. Don’t expect people to make the moves you want them to make, expect them to make the moves that they think are in their own best interest

That latter point is critical, especially as we now have rates at 0% in Europe, Japan, and the U.S.  We have QE programs in the U.K., the U.S., and Japan.  We have who knows what in China.  But each of these actions is causing other actions that may actually be the reason nothing seems to be working as well as it should in theory.

Do companies and executives really respond to QE the way the models predict or is their reaction different?  Maybe their reaction produces a better outcome for the company than the reaction the central bankers want and need out of those executives?

Too much of policy seems to assume certain moves will be made by other players when it is far from clear that those moves are either optimal for those other participants from their overall perspective.

As our balance sheet grows, as we create negative real rates, are we really sure we aren’t doing more harm than good, and what will the world look like 10 moves from now, or 50 moves or 100 moves?

Sadly, I don’t think anyone honestly knows.

What Does this Mean?

Mostly it lets me get this off my chest.  Somehow this topic has been bothering me a lot lately so I feel better having written about it.

But on a serious note, I think it is another reason to scale back positions.  Liquidity already seems abysmal, and this is a market largely supported by the faith that central bankers can continue to support it.  It is circular because the central bankers do keep getting forced to support it.  The longer this goes on, the greater the risk that we find that there is a problem, and that this “circularity” has been distorting values to the detriment of the economy and that the market loses this crucial element of support.

I find more and more people questioning the usefulness of central bank policy.  While I can see that the most likely path is a continued grind higher/tighter/better, it seems to me that there is growing doubt that the policies are working and any shift away from a full on love affair with central bankers is likely to be disruptive in a negative way.  I still think that is a low probability event, but that risk is growing and at this stage of the year, with so little liquidity, keeping risk low and even slightly bearish now is the right trade.


via Zero Hedge Tyler Durden

Leave a Reply

Your email address will not be published.