Learning to Trade with Robots
When you make a trading decision, you might assume that you are betting against another human who has done fundamental research and is willing to take the other side of your trade.
That view may be reassuring, but it’s about 20 years out of date.
Today, trading does not involve human against human. It involves human against robot, or, more likely, robot against robot.
So, the key to trading today is not outthinking another human, but outthinking a robot.
That may sound daunting, but it’s not.
The truth is that most robots are fairly dumb when it comes to trading decisions.
To realise profits you need to get your fundamental analysis right, but you also need to understand how the robots are programmed.
Once you have your view and you grasp the robots’ limitations, you’re in a good position to beat the robots at their own game…
Person to person trading
Investors may recall the markets of the 1970s and 1980s in which an individual or institution placed an order with a broker.
That order was then relayed to the floor of the New York Stock Exchange or to the Nasdaq stock market.
Specialists made a market in individual stocks. They matched orders with a view to maintaining an orderly market.
The specialists were also capital committers who would buy when most wanted to sell or sell when most wanted to buy.
These market-making obligations were the price that specialists paid for the privilege of seeing the back-of-the-book orders and knowing where buy or sell support existed.
On the whole, this system of price discovery at the point of sale worked well and served the interests of investors, specialists, and issuers.
By the 1990s, the order matching process was done mostly by computers and the role of the specialists as market-makers was greatly diminished.
Still, trading mostly consisted of individuals and institutions taking competing views against each other, even if the orders were matched by a computer.
The computer was a facilitator, but humans still called the shots….
Robot taught to trade in a range
By the early 2000s and increasingly in recent years, robots are the real decision makers. This is not just a matter of automated order matching. They are programmed to make buy and sell decisions on their own with no human intervention.
The robots do this using algorithms, massive data inputs, and artificial intelligence.
They are capable of reading financial reports, speeches, news headlines, press releases, technical reports, and literally millions of other sources, at speeds much greater than any team of humans could ever accomplish.
These inputs are processed using keywords and phrases such as ‘rate increase’ in a Fed statement or ‘drone attack’ in a Middle East breaking news feed.
The robot is taught to instantaneously buy or sell whenever a bullish or bearish signal is detected in the source material.
Robots are also taught to trade within ranges.
This is the 21st century version of the classic ‘buy low, sell high’ formula.
Robots will automatically buy oil when it dips below US$50.00 per barrel and automatically sell when it nears US$70.00 per barrel.
This accounts in part for oil trading in a narrow range recently.
(The same range-bound algorithms apply to stocks, with the robots programmed to sell when the Dow Jones hits 27,000 and buy when it hits 25,000.)
Humans cannot possibly absorb as much data or trade as quickly as robots.
Fundamental analysis trumps robo data
Does this mean the man-versus-robot race is a lost cause? Not at all.
The key is to understand how the robots are programmed, stick to your own fundamental analysis, and beat the robots at their own game by anticipating pre-programmed turning points as well as those inflection points where the robots will fail to anticipate a breakout from the trading range.
It’s not that hard.
Right now the fundamental drivers of oil prices are the collapse of Venezuela, sanctions on Iran, central bank rate cuts and stockpiling oil in anticipation of supply disruptions.
With geopolitical tensions and inflation expectations on the rise, what are the prospects for oil prices in the months ahead?
My analysis suggests that that oil prices are ready for a breakout to the upside after a period of consolidation.
The chart below shows the ICE Brent Crude Energy Future contract price (LCOC1) from 30 September 2014 to 30 September 2019.
The 30 September 2014 price of US$95 per barrel, was in the midst of the oil price collapse that started in mid-2014 (oil was at US$114 per barrel in June 2014 when the collapse began). Oil completed its plunge on January 20 2016, when oil briefly touched US$27 per barrel.
The Brent Crude Energy Future: 2014–19
Source: Thomson Reuters Eikon
Notwithstanding the volatility of the 2014–16 price collapse, oil has since traded in a relatively narrow range between US$40 and US$90 per barrel.
Most of the trading has been confined to an even narrower range of US$50–70 per barrel.
Right now oil is trading around US$60 per barrel, which is close to the middle of both the broad and narrow ranges.
Oil has also traced out a pattern of lower highs and higher lows since 2018, indicating a consolidation possibly leading to a breakout in either direction.
Humans still have an edge
My analysis is that oil is on the verge of a breakout to the upside.
This is based on a number of factors.
Central banks around the world are in a race to lower rates in an effort to counteract slowing global growth and recent disinflationary trends.
Rate cuts will do little to stimulate growth, but they can change inflationary expectations.
Monetary policy works with a lag.
The rate cuts occurring now in Europe, China, Australia and the US will produce at least mild inflation in 2020.
That will give a boost to oil prices despite slower growth.
The other factor pointing to higher oil prices is the geopolitical situation. Iranian oil exports have been subjected to sanctions by the US.
Other US sanctions on Iran under a campaign of ‘maximum pressure’ have created currency devaluation, inflation and social unrest.
The Iranians cannot beat the US in financial warfare, but they can retaliate with conventional warfare and terror; including drone and cruise missile attacks aimed at the Saudi Arabian oil processing facilities.
Iran has also disrupted tanker traffic in the Persian Gulf near the Strait of Hormuz.
If Saudi Arabia or the US retaliates for the Iranian attacks, the escalation will drive oil prices much higher.
If they do not retaliate, Iran will continue its provocations, which will have the same effect on oil prices. In either scenario, geopolitical tensions will increase and the result will be higher oil prices.
Demand for oil, and therefore higher oil prices, will also be driven by the need for stockpiling. China and Japan produce relatively little oil and are heavily dependent on Saudi Arabian oil to run their economies.
Any disruption in Saudi shipments to these countries would be extremely damaging.
Both countries have begun to stockpile oil reserves as a counterweight to possible supply disruption in the near future.
This is another floor under oil prices and creates an asymmetric trade where prices have the potential to spike higher, but are unlikely to fall significantly.
Robots are good at staying in trading ranges, but they are not good at interpreting geopolitical strategies or understanding inflationary expectations.
This is where humans still have the edge.
By understanding how robots react to news and by anticipating factors that they fail to process correctly, the nimble trader can legally front-run robots and earn huge profits from trends that good analysts can understand, but robots cannot.
All the best,