This blog and the other websites controlled by Greenteeth Digital Publishing have always warned against the tendency to be fooled by the hyperbole that is attached to concepts like Artifical Intelligence, Machine Learning and Algorithms.
“Algorithms will always make better decisions that humans because they are not hampered by emotions,” the people who long for they day when all humans will have processors implanted in our brains so our every action can be controlled by that bunch of fascist nerds who run Google. These people are not human enough to understand emotions are a vital part of decision making just as undefinable qualities humans possess are what propelled us to the top of the food chain.
One of the areas in which we are told machines have already surpassed humans is trading in the financial markets. Computer Algorithms can trade shared many times in a minute, which has been responsible for the crazy fluctuations we have seen over the past few years in stock markets. Computers, algorithms and mathematical models of various market sectors look at the numbers but do not see the bigger picture.
“Governments to ban petroleum fuelled vehicles by 2050 scream the headlines.”
“Buy Tesla, buy NIO, Buy Frarday Future the algorithms tell the automated trading systems seeing the likelyhood that these stocks will rocket in value. Go long on cobalt says the wise human trader, seeing after in depth research that sales of EVs in nations which have cut subsidies on green vehicles have collapsed, but calculation that despite technical and practical problems not being talked about (if you live in mountainous country, forget that Tesla, they’re rubbish at getting up hills,) EVs have a future as city cars and short range delivery vehicles. Cobalt is an essential component of lithium – ion battery packs, there are not enough cobalt reserves to make battery packs for a tenth of the number of vehicles now on the world’s roads and lithium is not exactly plentiful when we consider the amount needed to provide batteries for 1.4 billion vehicles (the estimated number on the road in 2017.
Logically, when news of a cut in subsidies, or the threat posed by China controlling 90% of the world’s cobalt reserves his the financial news feeds, algorithms will pick it up, stocks are going to crash in price, but demand for cobalt will still be there.
Rather belatedly (it’s algorithms spooked by the massive losses suffered in global markets towards the end of 2017 perhaps, )it seems the wall Street Journal has picked up on this flaw in algorithm driven trading.
Since about two years after the last great crash in 2008, markets rose in what appeared to be an irreversible trend, driven by the $£€ trillions of Quantitative Easing cash central banks [umped into the major economies. Algorithms were programmed to ‘buy the dip’ while frontrunning (an illegal insider trading practice,) every buy , virtually nobody – except for a few “fringe”, “fake news” blogs – complained about the threat posed by algo trading and thee accompanying deterioration in market stability.
Algorithm – driven trading enables trades to be conducted so quickly it is almost impossible to police this practice. A few financial pundits and bloggers who tried to warn that the practice would artificially inflate share prices were dismissed as fake news merchants as the official line that the economies of the liberal democracies were first recovering strongly and then forging ahead as demonstrated by competely false measures like the Dow Jones Industrial Index and The Financial Times 100 share index.
In December 2018 it looks as if the financial centres are entering their first ‘bear’ market since the 2008 crash, which those with long memories will recall was brought about by banks’ and investment funds’ exposure to dodgy debt derivatives (toxic debt,) and now as then politicians, media and regulators launched witch hunts as they sought scapegoats (anyone but the guilty banks of course,) and tried to deflect blame from the real culprit, then the irresponsible risks taken by traders looking for quick profit, this time, in their quest to help commercial banks make risk free profits the central banks, supported by governments, have pushed the cause of “computerized trading.”
An article on the front page of The Wall Street Journal, 26 December 2018, titled “Behind the Market Swoon: The Herdlike Behavior of Computerized Trading”, features contributions from a number of WSJ writers who collectively opine that “behind the broad, swift market slide of 2018 is an underlying new reality: Roughly 85% of all trading is on autopilot—controlled by machines, models, or passive investing formulas, creating an unprecedented trading herd that moves in unison and is blazingly fast.”
Absolutely true, furthermore a majority of the “autopilot” trading is also on the upside, that is on the assumption that a stock already trading at its highest valuation ever will go even higher. Strangely government finance chiefs, bankers and financial journalists did not see fit to mention this for several years. Of course, to new media readers, the story is all too familiar: after all we bloggers and citizen journalists have covered all of this not just on occasions when markets dipped sharply lower, but more importantly in the times when on an almost daily basis they soared to new, ever crazier heights, making an eventual crash inevitable.
From the WSJ article:
Today, quantitative hedge funds, or those that rely on computer models rather than research and intuition, account for 28.7% of trading in the stock market, according to data from Tabb Group–a share that’s more than doubled since 2013. They now trade more than retail investors, and everyone else.
Add to that passive funds, index investors, high-frequency traders, market makers, and others who aren’t buying because they have a fundamental view of a company’s prospects, and you get to around 85% of trading volume, according to Marko Kolanovic of JP Morgan.
Deutsche Bank argues that momentum has emerged as the most important force in markets, something we have claimed for years:
However, one key reason why trading has become so complicated for most, and certainly the algos, is that there is currently virtually no momentum in the market – with the MTUM ETF which tracks momentum stocks having its worst month and quarter since its 2013 inception – results in making any attempt to piggyback on the market a money-losing trade.
Bad news for those whose savings or pension pots are invested in index tracking funds. But what does it mean in plain English. Let’s turn to that WSJ article again and some quotes from traders who after a week of crazy fluctuations are suddenly very concerned about maret conditions:
Boaz Weinstein, founder of credit hedge fund Saba Capital Management LP, said the market had been underpricing uncertainty. Now it’s taking into account political issues “at the same time as the Fed is hiking, the economy is slowing, and a lot of people are feeling like the best days for markets are over,” he said.
Mr. Weinstein says there are dangers building in the junk-bond market. One worry, he says, is that so many junk bonds—he estimated about 40%—are held by mutual funds or exchange-traded funds that allow their investors to sell any day they like, even though bonds inside the funds are hard to sell.
When enough investors want to cash out, such a fund has to start selling bonds. But without much liquidity, finding buyers could be hard.
A selloff could start simply, he said. “It has its own gravity.”
The punchline to all this is delivered by COOPERMAN [Sorry US reders, that joke will only be accessible to UK readers of a certain age 🙂 ] :
“Electronic traders are wreaking havoc in the markets,” says Leon Cooperman, the billionaire stock picker who founded hedge fund Omega Advisors.
There is much more in the full WSJ article, which also addresses a collapse in market liquidity, the a knock on effect from equity markets to the credit market. If that happens, and it seems to be happening already, the stage it set for an almighty crash, but let’s face it, afall from around 26000 to 21600 between December 1 and close of business on December 24, followed by a 1100 point rally on 26 December is not sane trading, unless we accept that the behaviour of the electronic herd is “the new sanity.”
Pundits were not talking about algorithm driven, high frequency EFT trading “wreaking havoc in the markets” when the indexes were soaring ever higher. 2018 has been a terrible year for financial traders, the wild fluctuations have become more exaggerated than ever while public trust has fallen below decimal zero and is heading from absolute zero. This maybe explains why the WSJ article reports on the previously unmentioned aspect of what the force that propelled markets to such unrealistic heights. For around a decade now High Frequency Trading by computer, algorithm driven decision making and various other computer based trading technologies including the ridiculously misnamed Artificial Intelligence Systems that so many people not old enough to understand how computers actually work place such great faith in, have been overruling the professional traders most vital tool, gut instinct. Computers and their algorithms and Artificial Intelligence only do what their human programmers, who are rarely people who have ever been equity, currency or commodity traders, instruct them to. Thus they simply accentuate momentum either up, or down by frontrunning the orders from investors.
HFTs can only frontrun the flow of orders,to elaborate on the brief explanation in the opening paragraphs, looking at figures on financial news feeds only visible on the deep web it seems that after the massive sell off in December fund managers needed to tie up cash over the quarter end (I’m not going to go into why, I’m an IT systems person not a fund manager,) and placed orders for around $60 billon-worth of equities to bed-and-breakfast their cash for tax reasons. And the algorithms, having noted the sell off, followed their programmed instruction to “buy the dip” to the tune of that $60 billion.
In the process, instead of adding liquidity to the market, the HFTs sucked liquidity out of it. It’s good that mainstream media is finally reporting on the problem which will ultimately kick off the next big crash and possibly bring down the corporatist economy, i.e. the takeover of the market by computerized trading.
– a crash which, however, will only be made possible by the Fed blowing the biggest asset bubble in history to monstrous proportions, something which the WSJ article does at least acknowledge in its final paragraph:
“It’s not just about the equity market throwing a temper tantrum, it’s far deeper than that,” said David Rosenberg, chief economist at Gluskin Sheff & Associates in Toronto. “This is a much broader global liquidity story.”
Encouraged by signs of economic strengthening, the Fed has been gradually raising interest rates from rock-bottom levels and selling back the trillions of dollars in bonds it bought in the postcrisis years. The central bank says the roll-back of stimulus is smooth. Others aren’t so sure what comes next. There has never been such a huge stimulus, and one has never before been unraveled.
Some believe there’s a hidden risk in debt that consumers and companies took on when borrowing was inexpensive. The Fed’s campaigns were “fundamentally designed to encourage corporate America to lever up, which makes them more vulnerable to rising borrowing costs,” said Scott Minerd, chief investment officer at Guggenheim Partners. “The reversing of the process is actually more powerful,” he said.
Read the full WSJ article here.