Richard Craib a 30-year-old mathematician who runs a
hedge fund powered by artificial intelligence is Jack
Bogle-meets-Edward Snowden. Craib wants to democratize hedge
funds, and ultimately the asset management industry entirely,
in a way that would be recognizable to both men. Like Bogle,
founder of Vanguard Group and a lifelong missionary for
low-cost index funds, Craib rails against the cut Wall Street
takes out of investors returns. He thinks AI will lower
the cost of fund management, producing savings that can be
passed on to institutional and retail investors. Like Snowden,
Craib wants to shake the establishment to its core.
The very deeply strange thing about the asset
management industry is that no one is winning, says
Craib, a South African who favors black hoodies and started his
technology company, Numerai, in 2015. Thats why I
think there is an appetite for different ideas, including more
reasonably paid employees. The cost of our capital structure is
much lower, so we can charge much lower fees, he adds.
Numerais seed money came from First Round Capital, which
also seeded Uber, with later funding coming from Union Square
Ventures, Paul Tudor Jones, and others.
Quantitative investment firms or quants,
short for both the companies and the geeks working at them
use people to build systematic strategies. Numerai
intends to teach machines how to do that.
Industries from health care and finance to education are
grappling with AI, and the machine learning and other advanced
computing techniques behind it. The promise of autonomous cars,
which is shaking up every sector related to the automobile,
relies on AI. In March 2016, Googles DeepMind AI project
surpassed the best human mind at a game of Go, the ancient
Chinese board game in which possible moves outnumber atoms in
the universe. AlphaGo learned to recognize patterns as it went,
and beat Lee Sedol, the games world champion.
AlphaGos 37th move was one that no Go player had ever
seen before. The program learned the game through human-like
reinforcement learning, showed creativity in its moves, and
demonstrated for the first time the power of deep learning and
neural networks to create new insights and threats
for businesses and all aspects of human life.
Almost every asset manager is racing to hire people with AI
experience, and demand is especially acute from quants that
already have armies of Ph.D.s. Despite hiring high-tech talent,
these incumbents like their hugely profitable industry just as
it is. As hedge funds teach computers to buy winning
investments and some have already they tend to
package high-priced funds for investors in the same way hedge
funds always have.
Craib and a growing number of entrepreneurs schooled in AI
see a model ripe for disruption. After all, they have no legacy
businesses to protect. Experts anticipate that these
entrepreneurs, armed with computers taught to do the work of
expensive and moody humans, will first make inroads with a
lower-cost product covering the most efficient parts of the
market. Large-cap equities managers, watch out.
Investors always ask Renee Yao how she can compete with
entrenched quant giants like Two Sigma and WorldQuant. In 2015,
Yao founded Neo Ivy Capital, a statistical arbitrage hedge fund
that uses neural networks and deep learning, after running a
portfolio for one of those giants. One of Neo Ivys
original backers was Millennium, the firm WorldQuant spun out
We have a completely different business model,
explains Yao, whom Citadel hired out of Columbia University
before she could complete her math and statistics Ph.D.
Firms like WorldQuant hire researchers all over the
world, then use them to generate alphas. But were
researching neural network algorithms so those algos will train
the computer and the machine will find the alphas for us.
People need to rest and cant work 24 hours a day, seven
days a week. A computer can. Computers can work, generating
ideas for us, nonstop. This efficiency and lower cost
infrastructure only matter to investors if firms like Neo Ivy
offer their products at a lower cost. Although she declines to
detail planned fees, she says they will be significantly lower
than the 2-and-20 structure that leading traditional hedge
funds still charge.
Consultants arent counting out the entrenched quants
yet. Hedge funds and other active managers are grappling
with the concept of democratization, says Ben Sheridan,
partner at the Boston Consulting Groups San Francisco
office. Theyre exploring ideas on how to deliver
alpha at a lower cost, with fewer heads, in a machine-driven
way. But the changes existing managers are willing to
make are incremental, not sweeping. Sheridan expects the age of
AI to deeply change investing, but just how and
benefiting whom remains an open question.
This is an excerpt. To read the entire piece, click here.