Capital Markets – Calm before the storm

A team of physicists, natural scientists, IT and psychologists from University of Miami developed a model by which significant capital market movements, including the rare stock market crashes can be   prognosticated. The leader of the scientific team, Dr. Neil Johnson, head of the Department of Physics, former professor at Oxford, whom publications were cited many times by the Wall Street Journal. According to the model, the current situation is similar to the one that preludes major market upheavals.

The different scientific experts as a team profiled – through physical, mathematical, political, psychological analysis – many former crisis situations and other major market changes to be able to tell the human and artificial (robotic) responses to expected and unexpected external effects and their highly probable consequences more precisely by their model.

Péter Zentai: The model developed by your research team, made possible the capital markets to be well understood; the way millions of market agents will probably behave within a set timeline in a wide range of situations. Hence, forecasts for exceptional crises can become considerably precise… Does your model indicate anything significant referring to the recent global market?

Neil Johnson: Simplified, it indicates: the situation we are currently in, is essentially a calm before the storm phase.
Paradoxically, the main consequence is that there is no consequence – fundamentally speaking!
Leastwise, this features the capital market in recent months. The low volatility period has been unusually prolonged. However, storm clouds are gathering above our heads at the same time: for instance the sharp fall of exchange rates two-three weeks before. We must note that one of the major features of the current situation is that a wide variety of news (political, natural or medical) or unfounded rumours as well – even if they have minor or only local effect – are able to make unexpected changes – if not permanently, but distinctively – in the stock exchange rates, foreign- and raw material markets. The rumours of an American doctor having symptoms of Ebola while travelling on the subway in New York, resulted a sudden fall of exchange rates in Europe.

Why does the endeavour to understand this actual case and those other observations that say rumours and insignificant news as well are able to affect the prices require natural scientific studies? This is psychology. On the other hand, you are physicist and there are many mathematicians, physicists and natural scientists.

The fundamental difference of this model from earlier ‘market-monitoring’ models is that it integrates results from different scientific fields. We have been following the major market movements since 2006 – by not only our physicists, mathematicians, but our IT, economic- and stock market historians, investors and psychologists as well. We profiled carefully hundreds of relative or outstanding movements, including the outbreak and cease of the great credit crisis: we modelled the affecting human and the lately more and more influential mechanical (robotic) behaviours and the social, political, economic and financial circumstances as well.
In the last eight years, we lived through and processed situations which resemble the current one: when the volatility was permanently low, the human market-agents are struggling to find their way, while the ultrafast trading computers that can react within a fraction of a second are becoming more and more active.
These are the typical situations – in every case we studied – that came before great storms, like the great crisis between 2007 and 2009.
According to the conclusion of our model, high volatility is typical of markets when their agents are primarily hanging on real, economic and financial fundaments.
Low volatility has always been typical of periods that react to news and rumours without economic base. These times, people just loiter among news and rumours.
Usually, such events put an end to periods like the current one and initiate massive changes that previously were looked over or were not even registered.
For instance, on a plane, before taking off, we decide we are going to sleep, watch a movie in the next hour or write the presentation for tomorrow, so in any way, our journey will be effective.
However, a baby starts to cry behind us. At first, we think ‘the crying is going to stop eventually’, but after a couple of minutes we starting to get worried: ‘what if it isn’t?’. In the end, the baby cried during the whole flight: we could not sleep, work, or watch a movie either. An unforeseen condition (the crying of the baby) ruined every of our plans.
As for the current market situation: its prolongation keeps increasing the probability of major changes. The investor strategies and behaviours are becoming uniformed, because in such situation nobody can set a major goal for the close future; almost every person is mindlessly following protocols and copying others. While, at the same time, everyone wants to gain more yield in a typically yield-poor environment. This way, more people start to act independently that is unpredictable. They are feeling tempted and the overcomer ‘hunting instinct’ results in that people start to think they are smarter than the average. They are starting to step out of the boundaries of fair play in the hope of a greater yield, in other words, they are trying to outsmart others.
However, there is no way, people will not start to follow, especially in times like this, a behavioural or trading model that worked once for someone.

Regarding the effect of a crying baby on board – this metaphor can be applied to politics. Strong Arabian regimes were overthrown in the Middle-East as a result of student protests in Tunisia…even if their demands did not have world-changing intentions.

Moreover, a bigger, world-politically significant shift was triggered – indirectly – by demonstrations in the centre of Kiev which were first formed spontaneously, then more and more systematically against certain actions of the head of state. Many historical events support that street protests – that are originally formed against more minor issues – can disturb a country’s internal order or overthrow governments. They also support that even wars can break out because of seemingly insignificant arguments which could be resolved via communication.

However, the productive dialogue requires not only two significant ‘players’ but also that those two to be on a par with each other. In the world dominated by the Soviet Union and the United States, the outbreak of a war was prevented by parity.

A balance can only be formed by two major powers, indeed. However, this world can never return anymore, because of the increasing rate of information and technological developments which are accessible by everyone. On the contrary, the number of major powers is growing tremendously. Such factors are coming and can come to the front that not only have not been real ‘players’ but have not even existed before. For instance, the IS (Islamic State). The growing number of the states of the European Union is weakening the position of the formerly absolute powers, because the possibility of an uprising of smaller members has unpredictable consequences.

I assume, this is valid especially for capital markets. Along with that, the major and minor powers as well, just like car drivers, want to reach their goals without any casualties. It is in no one’s interest to start a war, because it comes with fatalities and requires great sacrifices.

Still, there are many accidents. Wars break out. Market crises happen. And the more people go for the same goals, for example, to generate more profit on the exchange market, the more will suffer deficit. We are modelling this continuously and it getting to seem obvious that market agents are basically repeating themselves. In details: they act according to some templates in response to certain challenges. It is possible to estimate the number of challenges which aid significant market changes. The model which integrates historical, physical, mathematical and psychological patterns is starting to give better answers when forecasting potential crises.

I don’t completely understand. If it is true, that eighty percent of the transactions are carried out by intelligent and ultrafast computers in the American markets and it is more than forty percent in Europe, then every historical and psychological model must be invalid.

This is the reason why both results of human and robotic behaviour studies are integrated in our model. We can say that there is an ongoing global war between rival trading computers and their programmes in relation to speed and the amount of actions done in the capital market, mostly in foreign exchange market and arbitrage operations. We documented almost twenty thousand cases in the last eight years where there were a clear correlation between price fluctuations and the expanding robotic traders.
Far be it from me to say, the ultrafast, algorithm based computers are fundamentally contributed to the development of major financial and credit crises but one of our related study is scientifically proved. At the beginning of our modelling, in 2006, we saw that computers are continuously transact sales in a well-defined sector and keep decreasing the share prices of the sector’s companies and corporations – in an increasing rate. The number of jumping outs from these shares exceeded the jumping ins and in the end there were nobody joining but only selling. Companies were targeted such as Lehman Brothers or J.P Morgan and the AIG. The operations of computers contributed to the sector’s extraordinary fall in prices.

What is the moral of the story?

Computers – at least in operational level – recognised anomalies of financial companies earlier than humans. However, computers after all are acting according to human-made programmes that are easier to keep track of than human behaviour. In other words, computers are more predictable ‘beings’ than humans are.
The more ground computer technology gains in the stock market, the more reliable our model and its prediction become – for example when forecasting major stock market falls.
It is also true, algorithms are working only in short terms, that in theory – or once experienced in practice when four years ago, the New York Stock Exchange fell ten percent within minutes without any rational reason – can generate selling (or buying) panic with unbelievable speed, or can even lead to market crash.