You are currently viewing our desktop site, do you want to download our app instead?
Moneyweb Android App Moneyweb iOS App Moneyweb Mobile Web App

In markets gone mad, investors find rare comfort in data science

A program that captures market fears is the next big thing.
The US-China trade war, Brexit and political tensions in Hong Kong are just some of the points of global uncertainty that continue to cause market turmoil. Image: Bloomberg

The Phillips Curve is in doubt, the bond market is distorted and a tweet from President Donald Trump can shift the trajectory of global markets in a matter of seconds.

In a world where traditional touchstones of fund strategies are being challenged by unprecedented economic-policy uncertainty, investors are seeking information not tapped before or better ways to sift through it. Some are stepping up the use of machine-learning to capture market sentiment on everything from the trade war to recessions. Whether these strategies are more effective than conventional methods remains to be seen, but they’re already paying off for some funds.

In a Bleecker Street loft in downtown Manhattan, Vasant Dhar, the founder of a $400 million hedge fund and a pioneer of AI investing, finds his computer-driven trading is just the thing for the Trump era. He uses a program that captures not just securities prices, economic data and news sentiment but also market fears — studying patterns of volatility. In Boston, Eaton Vance has a four-person data science team studying anonymous credit-card spending information, customer sentiment from social media and ETF flows on top of the fundamental work of its equity portfolio managers.

“At times, the market has been behaving like a naive schizophrenic that puts all faith in only the most recent trade tweet,” said Eddie Perkin, chief equity investment officer at Eaton Vance. “As humans, we know we have intellectual biases that can lead us to flawed decision-making such as reacting to false signals from the market. We have built our investment processes to address this and help keep us out of trouble.” The approach has helped Eaton Vance’s Small-Cap Fund gain 22% year-to-date, almost double the Russell 2000 Index’s 11.6% increase.

A lot that markets relied on in the past has been turned on its head. The challenge to the Phillips curve — the notion of a trade-off between low unemployment and higher inflation — is clear, with Federal Reserve Chairman Jerome Powell saying in Zurich last week that it’s “not what it was.” Many say the bond yield curve, once a barometer for economic outlook, has lost its predictive edge after central banks’ unprecedented monetary stimulus. A US president who makes major policy announcements through tweets has many funds scrambling to account for the heightened turmoil.

The US-China trade war, Brexit and political tensions in Hong Kong are just some of the points of global uncertainty that could cause further market turmoil. Trillions in government-debt purchases by central banks have driven the term premium — or the extra compensation investors required to own long maturities rather than roll over a shorter-dated obligation — to record lows. More than $15 trillion in global bonds now have yields below zero and the voracious search for safety has sent Treasury yields plunging.

The level of global policy uncertainty is not only palpable to investors but also measurable. A global composite index of economic-policy uncertainty of 20 countries hovers near a record high reached in December.

Against that backdrop, analysts at JPMorgan Chase & Co have created an index to gauge the impact of Trump’s tweets on US interest rates, which they say is on the rise. The ‘Volfefe Index’, named after Trump’s mysterious ‘covfefe’ tweet, suggests that the president’s posts are having a statistically significant impact on Treasury yields. Quants at Citigroup are sending currency-market clients notes analysing Trump’s missives on Twitter.


“I don’t think a lot of people are systemically hedging their trade-factors exposures,” said Basil Qunibi at Atom Investors, a hedge fund in Austin, Texas, he founded. “That explains the fact that when there is a tweet or a significant change in tariffs or the trade war overall, people tend to react in a pretty aggressive way.”

The Dow Jones Industrial Average slid as much as 600 points on August 23 amid a Trump tirade after China imposed new tariffs on American goods. Trump followed with tariffs on roughly $110 billion in Chinese imports on September 1.

Trade woes have added headwinds to the Fed’s struggle to get inflation toward its 2% target even amid a robust US labour market, which includes an unemployment rate hovering around half-century lows. The European Central Bank is so concerned about its failure to lift inflation, it’s seen announcing a large stimulus package on Thursday.

For people like economist-turned-investor Stephen Jen, who’s the chief executive officer of Eurizon SLJ Capital, the market is right to pay attention to Trump’s tweets because the US president tends to do what he says. Jen’s more worried that central banks may have failed to catch up with the structural changes in inflation, which is increasingly being influenced by globalisation, technology, and demographics. That’s due in large in part to players like Amazon and Alibaba Group who source and deliver all over the world.

“Inflation is now a global variable, not a local variable, and it should be dealt with in a global context,” said Jen. “We are now left with a situation where central banks are still committed to inflation targeting they set up 20 years ago. Their navigation system may be outdated, and we might end up being an 18-wheeler that gets stuck in a small village.”

Falling inflation expectations, plunging bond rates and flatter yield curves all point to mounting doubts in financial markets over whether monetary policymakers have what it takes to reflate their economies and avert a global recession.

Uncertainty has become a global whack-a-mole game, with the temporary calm in one hot spot merely being surpassed by blow-ups in others.

As the geopolitical twists and turns and the streams of tweets from Trump roil markets, in his loft in Manhattan, Dhar, who’s also a professor of data science and artificial intelligence at New York University, is pleased that some years ago he put computers in full control of the Adaptive Quant Trading program that he developed for his hedge fund SCT Capital Management.

“You just can’t figure out what will happen with the trade war,” he said. “It doesn’t mean you’re not concerned, but you’re not chasing your own tail and wondering what to do next and what’s Trump going to say, what’s he going to tweet. That stuff factors into the fund’s positioning but indirectly through the patterns it sees in prices and volatilities.”

© 2019 Bloomberg L.P.


Sort by:
  • Oldest first
  • Newest first
  • Top voted

You must be signed in to comment.


Geez AI is an oversold hype, people looking for patterns but fooled by randomness.

There are some applications in bounded/stationary problems sure, however self driving cars for instance will result in many “unforeseen” or “1/ billion” events causing many deaths.

With the difficult tasks of forecasting standard stats models outperforms.

Casi, AI is a modern-day-word for voodoo!

I’m not sure there is enough historic data points for AI to be able to predict the future with reasonable accuracy, given the current environment (Brexit, Trump, China)

The bright side. Investors in history created the environment for hunting and better. Staying at home, Jan sailed this way on investment prayers. Zimbabwe, another story, had a funeral of the men who really outdid Verwoerd and Klerk in uplifting himself. With no investors around, this peace of planet, called Africa, should be paradise as seen from hopeful Malema eyes. Admiring the result of his hero. Setting the first step back to a world of happiness.

“There is nothing new in the world except the history you do not know.” -Harry Truman.

There is nothing new about the current monetary situation. We are merely at a certain point in the cycle. The names change, the technology changes, the currencies change, the names of the banks change, different economists invent new models, but the situation remains the same.

Kondratieff described the business cycle within the longer-term gold/fiat currency cycle. Once the connection to gold is severed, the trend is always towards more debasement, never towards a return to sound money. Debasement leads to the rise in assets prices in nominal terms. Rising asset prices lead to more taxes, happy savers, bigger government, more socialist programs and more devaluation.

You can merrily keep on running on this rat wheel, or you can decide to start pricing everything in terms of ounces of gold. Then you will see how everything stays the same.

I find myself agreeing with you on almost everything except your views on the gold standard. Yes, fiat isn’t great and the leverage involved in fractional reserve banking isn’t helping either. The Song dynasty, after all, invented paper money and hyperinflation in short succession.

However one has a choice: one has to choose between the gold standard and democracy. These two are mutually exclusive. Why? The moment one has a trade deficit, deflation starts and wages plummet, followed by either civil unrest or a democratic change in government. Countries shift their focus from any and all matters towards maintaining gold reserves. Why not run a surplus then? Because if everyone is running a surplus, no-one is. The gold standard leaves no way to move surpluses around, because the focus is now on maintaining reserves.

In fact, the Euro is for all practice and intent a gold-standard. It has all the hallmarks of gold. A country on the Euro has no control over ECB policy, cannot print money, cannot devaluate its currency, cannot inflate out of debt, can do nothing about interest rates and cannot control the money supply. These things seem good until you realise that the euro works only for surplus countries (Netherlands and Germany) and devastated the economies of deficit countries (the entire Med).

End of comments.





Follow us:

Search Articles:Advanced Search
Click a Company: