The oil price that had been going down at an
alarming rate in the last two weeks, got a catalyst on Wednesday morning out of
the blue, when a massive cargo ship ran aground in the Suez Canal in the middle
of a storm.
With a gigantic obstacle blocking the path in the
busy waterway, tankers have been spotted idling behind it, carrying millions of
barrels of crude oil; understandably, the fear of an imminent supply shortage
blew across the markets and the oil price went up.
The scenario, once again, proved the fact that the
sentiment is the king in the paradoxically fluctuating crude oil markets; it is
an unsurmountable challenge to the data analysts who work, day in, day out, in
search of the magic formula to predict the price accurately.
So far, this pursuit has been of no avail.
Last week, most pundits, who monitor every
conceivable factor that could determine the crude oil price, were in unison in
declaring the decline in oil price; that, however, dramatically changed
yesterday, when the news of the blockage of the Suez Canal started emerging.
In this context, until we come up with a mechanism
to analyse the prevailing sentiment and quantify it, despite being monumental,
the reliable prediction of oil price in a few days’ time, let alone in a month
or so ahead, will be just a futile exercise.
Unfortunately, even super-complex algorithms,
involving the Machine Learning, have made no difference either, despite
crunching enormous amount data at an unbelievable speed.
In short, the Holy Grail of price formula is still
as elusive as ever it has been, despite the amazing advances made in the Data
Science.
Perhaps, if we manage to analyse the universal
markets sentiment in an accurate manner, we may get closer to the magic formula
that could predict something reliable.
In order to become a part of this seemingly-losing
battle, I subjected an article on the Sues Canal blockade to a sentiment
analysis algorithm, as shown in the animation above, and it was pretty
negative; I repeated the experiment with a few more articles in the British
newspapers in order to spice it up statistically and sentimental factor was
fairly consistent; but the number was still fluctuating considerably.
All in all, the ball is now in data scientists’
court; they need to come up with algorithms to get the sentiment right – along with
other factors, of course - in order to make the customers, who rely on their
predictions to get the investors out of a sphere of perpetual nervousness.
