George Soros made an interesting point in his book '
The New Paradigm for Financial Markets: The Credit Crisis of 2008 and What It Means': financial markets are human!
He advocates for a new paradigm for financial markets--I agree that there is a need for a paradigm shift!
But THEN he also advocates for increased regulation of financial markets... The problem with that:
MOST REGULATORS DON'T HAVE A CLUE OF WHAT FINANCIAL MARKETS ARE OR HOW THEY FUNCTION.I believe what's needed at this stage is an overhaul of financial models; to 'make them human'. I made this point in issues 12 & 14 of
The Price Report:My contribution from Issue 12 of 'The Price Report', dated 18 September 2007:
Why Models are ugly--and how we can make them better looking
In this and future articles, I’m hoping to open up a debate among readers of
The Price Report. My thesis? In essence, that while there are endless financial models, probability analysis methods and other mathematical and statistical tools used to forecast market events: it is in the understanding of people and how they make decisions that real market understanding lies: figure these out and you’ll be able to predict market events with a frightening degree of accuracy!
Read any financial publication from the last six weeks or so, and there will be one phrase you are sure to encounter at least once an article if not once a paragraph. It is the US “sub-prime mortgage market”. The phrase is usually used within the context of a discussion on the financial turmoil that occurred in the global markets “because of the poor performance of the US sub prime mortgage market”. But was the poor performance of the sub prime mortgage sector the REAL reason why the global markets are experiencing such distress at the moment? And the real reason why the UK is seeing a run on a bank for the first time in 30 odd years?
It seems to me that the crisis is less to do with the ‘market’ than to do with the market players who somehow failed to foretell, and prepare for, the oncoming catastrophe, despite the fact that in hindsight it looks so very obvious.
So why did they fail? Because they made the wrong decisions. And why did they make the wrong decisions? Now we get to the heart of the matter. They did so because their decisions were based on inaccurate methods or on the employment of sub-optimal tools.
I firmly believe that algorithms, probability analysis and financial models are all deficient in that they give one-dimensional forecasts of the market. They neglect or under-emphasise the role of human-beings in the market-place and how different ‘buyers’ and different ‘sellers’ make different decisions (specifically buy and sell decisions) when confronted with the same environmental stimuli.
The market consists of ‘buyers’ and ‘sellers’ with different personality profiles and different motivations. These investors don’t all act in a uniform, ‘rational’, ‘profit-maximizing’ or ‘pain-avoiding’ manner – as most models assume.
The state of the market at any point in time, or rather the prevailing price of an asset in the market place at any point in time, is a function of the
human beings that populate markets and different human perceptions of their environment and events, human interactions, and differing human needs and wants.
...This means that there is a clear need – for financial models that are more accurate in predicting real events in the market place: financial models that factor in the human component of markets...
The most critical component of any market, is the component of human beings within it. If we posses greater understanding of our fellow investors, of how they make decisions– often under conditions of great stress and inevitably somewhat limited understanding something which means they aren’t always rational - and how they interact with each other in the marketplace, we’ll be able to better understand market dynamics and predict market events more accurately.
So how can we respond to the shortcomings in our methods? By merging of the disciplines of finance, economics, mathematics, statistics, marketing, sociology and psychology. Such a synthesis across different disciplines would create financial models that incorporate human subjectivity and variety, how we differ as investors and how we interact within markets. This, in turn, would determine anticipated events in the market place. Such models would be more complete, and will give more accurate predictions of events that have a likelihood of unfolding in the markets. I’m not suggesting this is going to be easy, but in the next issue of
The Price Report I’ll explain how I intend to make a start on creating a financial model that incorporates a human component.
My contribution from Issue 14 of 'The Price Report', dated 16 October 2007:
A financial model that accounts for the human factor
In Issue 12 of
The Price Report I discussed:
1. How financial models, algorithms and statistical tools employed for market analysis give inaccurate or one-dimensional forecasts of future prices and future market events, because they neglect or under-emphasise the most important and basic aspect of any market—which is the human beings that populate the market.
2. How there is now a need for financial models that factor in human variety and a need for financial tools, or methods of analysis that acknowledge that different market participants have different motivations and personality profiles.
3. That all market participants may not act in the ‘uniform’, ‘rational’, ‘profit-maximising’ way that traditional models assume. Financial planners are conscious of human variety, differing motivations, different personality profiles and how these shape investment decisions made by their clients. Yet the majority of economic models and financial models still assume that human beings act in a ‘uniform’, ‘rational’, ‘profit maximising’ way, as homoeconomicus, even as all financial players acknowledge that there is a huge variety of motivational factors acting upon market participants. As individual investors, we defy crude labels.
4. That a synthesis across the disciplines of economics, finance, mathematics, statistics, marketing, sociology and psychology is needed to create financial models that are more accurate in predicting future market events and future prices.
In this piece I’ll discuss possible ways of constructing financial models – specifically asset pricing models – that factor in the human component of markets.
So how does one start the construction of an asset pricing model that factors in a human component?
I believe that the best starting point is the one I identified in my previous article: that the prevailing price of an asset in the market place – at any point in time – is a function of: different human needs; different human wants; different human perceptions of events in their environment; and different human interactions within their environment.
Mathematically this can be expressed as:
Pt a = f(Hn, Hw, Hp, Hi)
Where:
* Pt a – Represents the prevailing price of Asset ‘a’ in the market, at a specific point in time – denoted by the letter t, which could be measured in seconds, minutes, days, or any standard. Pt a is expressed in a standard unit of currency e.g. dollars, pounds, yen etc.
* Hn – Denotes human desires, up to and including point t in time; which are critical for survival. These are basic human needs according to
Maslow’s Hierarchy of Needs i.e. the first two levels of Maslow’s hierarchy of needs. Hn basically represents physiological ‘needs’ (e.g.food, shelter) and security ‘needs’ (e.g. safety and freedom from harm/threats).
* Hw – denotes human desires, up to and including point t in time, which are not critical for survival. These desires can be basically described as ‘human wants’. According to
Maslow’s Hierarchy of Needs, this includes the three top tier ‘needs’, namely, social ‘needs’ (e.g. approval, belonging, love), esteem ‘needs’ (e.g. recognition, self-esteem, recognition) and self-actualization
‘needs’ (e.g. self-discovery, achieving the best you can, or being the best person you can be)
* Hp – denotes human perceptions of events in their environment (both the micro and macro environment), up to and including point t in time. These influence human decisions within the market place, particularly ‘buy’, ‘sell’ or ‘hold’ decisions.
* Hi – denotes human interactions within their environment, up to and including point t in time. Human interactions can basically include all social exchanges, which occur among human beings within their environment.
Note: The above-illustrated function is not plot-able using conventional methods, as it has five variables – conventional geometrical dimensions accommodate a maximum of three variables, on the standard 3 dimensional x,y,z axes. This means that one would need a 5-dimensional ‘hyper-space’ to graph the above-mentioned function – which can ONLY be done using hardcore engineering/mathematical software!
…The easy part is done, now to the messy business…
According to July estimates of the World Population by the C.I.A World Fact-Sheet, the world has a total population of approximately 6.61 billion people. So if the prevailing price of an asset – at any point in time – is the function of different human needs; different human wants; different human perceptions of events in their environment and different human interactions within their environment, what does this mean?
Does it mean we have to find out the needs, wants, and perceptions of 6.61 billion unique people to be able to forecast the price of an asset using the function I introduced?
Luckily it does not.
According to Stuart Litchman, the author of ‘How to Make Lots of Money for Anything FAST’ and creator of the artificial intelligence system called Arintel, there are 12 subconscious ‘personality types’ that are invariant around the world.
Firstly, this means that we ONLY have to analyse 12 ‘personality types’ instead of analyzing 6.61 billion unique ‘personality types’. Why? Because if you met a random person on any street, in any part of the world, from any culture, you can be sure that their ‘sub-conscious personality type’ belongs to one of the 12 clusters that Litchman identified.
The only difference between cultures is the proportion of each cluster you’ll encounter.
In the next issue of The Price Report – assuming Tim invites me back – I’ll explain how I intend to incorporate the 12 unique sub-conscious personality types into the function I’ve just introduced.