We have reached the point where the study of psychology has invaded almost every discipline.
Sports teams employ in-house psychologists, governments consult behavioural science gurus,
and supermarkets borrow strategies from neuroscience to boost their sales.
Economics is no different, with behavioural finance steadily gaining in popularity. Academic
psychologists and economists have long argued over the merits of each discipline, a conflict which
must be swiftly resolved as both economics and psychology is needed to predict the path of asset
prices in circumstances – such as today – where productivity growth is stalled and where nominal
interest rates are trapped hard up against their zero lower bound.
First, a quick look at the two theories:
Classical economics has its roots in 18th and early 19th century thinkers like Adam Smith and
John Stuart Mill, who proposed solutions to Europe’s unstable protectionist and mercantilist
economies. They believed free trade and free competition between workers and businesses
would more efficiently allocate capital and labour, where Smith’s hypothetical butchers and
bakers would rationally maximise their own welfares by taking decisions in their own best
interests. In this way, they argued the economy would become self-regulating.
Classical economics ended around 1870 when western economies were radically changed by big
leaps in electricity, the internal combustion engine, domestic running water, telecommunications,
and petroleum refining. These inventions rapidly expanded the middle class and founded the
economic systems we know today.
While classical theory assumed that a product’s value is derived from cost plus labour,
neoclassical theory puts consumer perceptions of the value of a product or service centre stage.
These theories hold that consumers ‘rationally’ maximise their personal utility (personal welfare),
that competition leads to an efficient allocation of resources within an economy, and that supply
and demand dynamics will naturally underpin a stable market equilibrium.
Behaviouralists bring these background behavioural assumptions to the fore. They emphasise the
cognitive factors which prevent individuals from behaving rationally, arguing that humans have a
limited capacity to act logically and as such approach decision-making with a host of cognitive
biases, illusions, and heuristics. Because individuals operating within the economy are inherently
irrational and unpredictable, behaviouralists argue that markets are largely driven by
psychological forces.
Neoclassical economics has consequently come under scathing attack from behavioural
economics for its superficial treatment of human psychology. To claim that an economy is self-
regulating, we must assume that individuals operating within it behave like robots and be so
sensible and intelligent that they always make optimal decisions. While classical economics gives
us an idealised model of how the economy should behave, human beings are in reality not
hyperrational utility maximisers and make all sorts of questionable decisions.
These criticisms are in truth something of a strawman. Neoclassical economists never seriously
suggested that complex economic systems and markets can be perfectly efficient. Neither are
behaviouralists as heretical as they might like to think: we have known since the Tulip Mania Crisis
that human psychology is one of the horses pulling the economic chariot.
The six-million-dollar question for investors and economists today is whether developments in
psychology can be successfully applied to economics in circumstances where conventional
monetary tools no longer work. Psychologists similarly wonder whether developments in
economies trapped at the zero-lower-bound of rates have implications for the core of psychology.
If it is possible to establish verifiable principles of human behaviour, then they must surely have a
place in theories about human behaviour to which psychology aspires.
The question here is whether the two theories can be successfully reconciled.
They can, if we acknowledge that they are not two different theories but two levels within the
same theory. This is defined as a ‘two-level theory’, which exists both on a primary surface level
and a deeper hidden one, with the deeper level causing outcomes at the surface level and vice
versa. What defines two-level theories is that each level is related to one another and neither can
be fully explained without consideration of the other. Think of a plant: there might be a beautiful
flower above the earth (surface level), but that flower couldn’t exist if minerals weren’t being
sucked in beneath ground (deeper level).
While philosophy has long recognised two-level theories, arguably the most relevant and beautiful
two-level theory is Einstein’s General Theory of Relativity where surface effects are constrained by
two deep principles: the principle of relativity and the principle of the constancy of the speed of
light.
Such an approach holds the possibility of resolving the clash between economics and human
behaviour. Like this:
A two-level economic model recognises both a surface economic level and a lower psychological
level – i.e. explicitly recognises that classical and behavioural economics are opposite sides of the
same coin, meaning that the psychological elements are treated more rigorously than just
assumed hyperrationality.
The name of the financial game is to predict surface level effects such as economic parameters,
stock prices, etc. However, if we try to explain them in a zero-interest rate environment as though
they are not part of a two-level theory, then we will always miss the mark.
Classical economics focuses purely on the surface level, and in doing so trivialises the agency of
human behaviour. Conversely, behavioural economics focuses on the deeper psychological level
and neglect to consider the macroeconomic forces which govern the whole system.
Two-level models can resolve this conflict.
Economic/Asset Price Two Level model:
Take contemporary econometric models as an example: these models require mathematical
constraint afforded either through assumptions we hope turn out correct, or by historic data sets
we hope are representative of the future economy, or both.
First, the collapse in the natural rate of interest (r*) amongst other deep changes to post-2008
economies has rendered it perilous to assume that historic data sets are representative.
Second, many conventional assumptions have become unsafe. Conventional monetary theory
assumes that r* is sticky and remains in a 3-5% range, and that increased capital intensity will lead
to improved productivity, which will in turn average 2-3% in the longer-term. In terms of asset
price models, Black Scholes assumes that stock returns are lognormally distributed, the interest
rate term structure is flat, etc. None of these assumptions hold today.
The loss of relevant past data sets and underpinning assumptions washes away conventional
notions of a business cycle, debt sustainability trajectories, and creates profound problems for
financial models including Black Scholes, CAPM, and Markowitz.
By comparison, a two-level model can be constrained by using generalised principles acting within
the lower psychological layer as substitutes for the lost assumptions and past data calibrations.
Difficult as devising these new models may be, the alternative approach is simply not viable.
Turning the handle of broken models is always a poor solution as attested to by the litany of
contemporary forecasting failures spanning asset prices and economic variables.
For those who work in finance, there is no better argument as to why you need to integrate an
understanding of human psychology into your armoury, as attested to by the increasing number
of conventional forecasts which turn out to be superficial, post hoc, or plain wrong.
If we could understand the economic without the psychological, we wouldn’t need Behavioural
Finance, Neuroeconomics, and Investor Psychology; nobody would read Thinking, Fast and Slow
and Daniel Kahneman wouldn’t have won the Nobel prize in Economics (as a psychologist). More
than ever we need to educate ourselves on biases, cognitive illusions, and heuristics, and then
pull them to the forefront of our thinking.
Your investment performance will only ever be a result of one thing: the decisions you
make.
It would be convenient to pretend that we make decisions like rational, welfare-maximizing
robots. The trouble is that this blinds us to the investment fads, fashions, and manias powered by
deeper human emotions, and which central banks quite deliberately target in pursuit of their
given inflation and employment mandates.
Accepting economics as a two-level theory is to accept that you cannot work in finance today
without recognising that the economic and psychological are increasingly impossible to separate.
True, nobody has come close to following the footsteps of Albert Einstein in developing a general
theory of markets, but that is not to say early attempts are without merit – far from it. We can only
understand where something is heading if we understand where it comes from and something
about how it works. Contemporary two-level models can’t do that but can often discern enough
about the underlying principles to figure out whether a security price is under/overvalued, even if
we do not know by exactly how much.
Simply knowing this is of course highly useful!
Matthew S. Machin