The chief executive of a discretionary equity hedge fund might seem an odd choice to include on a panel at a quant conference. But G Squared Capital’s presence at Battle of the Quants in New York earlier this month illustrates just how much crossover is occurring between the two philosophical encampments of the investing world.
G Squared takes a hybrid quant-fundamental approach, using artificial intelligence algorithms to analyse 10,000 data points per company and make investment calls.
The firm is in the vanguard of a quant fund push to mimic their fundamental cousins – now that the quants have a mounting pile of new data to crunch and machine learning algorithms to help make sense of it.
Michael Graves, chief executive of global stat arb hedge fund Nebula Research, who spoke on the same panel, said quants are starting to work on the questions fundamental investors used to see as their own: how is Apple’s new launch going down, or have Tesla’s plans hit a speed bump?
Nebula, for example, now trades an equity long/short strategy in a systematic fashion. “We are mixing some of their techniques with our techniques and trading on different timescales,” Graves said. He hopes the explosion of alternative data and improvements in computer processing power means the firm can do it better.
Michael Recce, chief data strategist at fundamental asset manager Neuberger Berman, predicted the amalgamation of systematic and fundamental methods of investment “will be larger than the original quant revolution”.
Neuberger acquired quant research firm Breton Hill Capital in 2017 to deepen its capabilities in that area.
Recce described fundamental investors as having knowledge “an inch wide and a mile deep”. In other words, they know a huge amount about the future value of just a handful of companies.
Imagine if you are building artificial intelligence that acts like a discretionary trader. You get the alpha of the discretionary guy and the leverage of the quantMichael Recce, Neuberger Berman
They can generate a lot of alpha from those few select bets. But they are limited in their use of leverage by the high idiosyncratic risk in their portfolios.
Quants on the other hand, gather knowledge “an inch deep and a mile wide”. They do not know much about individual companies, but they rely on finding mispricing patterns that repeat across many stocks over time.
Although systematic strategies typically only have access to relatively little alpha, they can lever up to achieve comparable returns to discretionary managers, Recce said.
“But imagine if you are building artificial intelligence that acts like a discretionary trader. You get the alpha of the discretionary guy and the leverage of the quant.”
This is the sweet spot for quants – to include alternative and additional fundamental company data in their models. Recce insists machine learning analysis on alternative datasets will be an advantage to any quant firm looking for inefficiencies in the market.
He pointed to January 3 this year when Tim Cook announced that Apple had sold fewer phones than anticipated in China, wiping $75 billion off the company’s stock.
“Everyone was surprised, but all those phones have carriers, make calls and have apps. The data is absolutely in the world, but it is not in the market,” he added.
JP Morgan Asset Management and Golden Sachs Asset Management are among those to champion the use of algorithms to trawl though news and social media to understand what is happening at companies before anyone else.
“If you watch news and social media carefully and use algos to read it, then you can find out something is going to happen before everyone. Say Panasonic is going to opt out of investment in Tesla’s Gigafactory. You don’t know what the true value of Tesla is… but you know temporarily that Tesla stocks will go down so the race is on for who can trade on that the fastest,” Recce said.
The next big trend in quant, it seems, could be a return to fundamentals.