“Can big data actually be made useful? Can it help us develop better products?” The rhetorical questions are being asked by Michael Kollo, London-based deputy head of research at AXA Rosenberg Equities, a €18 billion quantitative investment shop.
The answer is not yet a clear ‘yes’, and proponents such as Kollo are exploring how valuable machine learning will be. “We’re experiencing a data renaissance, but it requires investment,” Kollo said.
The stakes are high: if the answer turns out to be ‘yes’, it means machine learning will do more than build a better mousetrap: it will completely change how quant firms make their buy and sell decisions.
Kollo sat down with DigFin to explain the thinking behind how AXA Rosenberg is harnessing machine learning to its quant investment business.
Kollo has worked as a quant at asset managers including BlackRock and Fidelity. They all have their own business models. AXA Rosenberg pursues factor-model investing, using big data sets and machine learning. He is also charged with promoting innovation within the larger AXA Investment Managers organization – “Data sets are becoming institutionalized,” he said – but quant businesses are more natural adopters of machine learning.
Projects so far include tracking sentiment via social media to develop real-time indicators for portfolio managers (starting with inflation, for the fixed-income team), developing credit-scoring models (in partnership with French fintech DreamQuark), and using machine learning to assist the sales team pitch new ideas to customers.
More fundamentally, Kollo believes machine learning will help asset managers rethink their production process, rather than just help them do the same thing more efficiently.
What the machine sees
Machine learning is a subset of A.I. The science of making computers intelligent relies on machine learning to train them into making autonomous decisions. Kollo’s brief primer serves to illustrate how this will change portfolio decisions.
He shows DigFin an image of the numeral 8. The goal is to get a computer to always recognize this numeral, even though people might write it in all sorts of ways. Even so, we will recognize a numeral 8 when we see it. A machine cannot intuitively grasp what makes an 8 an 8. So it has to be taught in another way. (The images Kollo showed actually come from a series of posts on Medium by an A.I. specialist, Adam Geitgey.)
A machine will read an image of the numeral as a series of spaces that are colored lightly or thickly, or not at all. It divides the image into many small blocks. The computer assigns each micro-block a number that corresponds with the level of darkness it sees there.
When this is reproduced as an image, you can still make out the numeral 8, as a series of bigger numbers where there is ‘ink’, and zeros or ones where there is empty space. A human looking at this first rendition sees a recognizable if abstract numeral 8 among a field of numbers.
This taking inputs – a series of numbers that correlate to the amount of ‘ink’ visible in an image – and turning it into a set of numbers creates what is called a vector.
But another step – another set of vectors – is required to understand what is happening. The abstract numeral 8 amid a field of numbers only looks like an 8 because the grids are arranged in the same shape as the original image. But the computer doesn’t want or need to see this in, say, size A4. Instead the numbers simply run on, in order, but without regard to shape. To the human eye, the numbers run on into gibberish. We no longer see the 8.
But the computer does, when it is programmed to. Its neural network reconceives this string of numbers – random to the human eye – as a probability: the probability that this vector means “8”. The process of turning a number into a probability is called ‘logistic transformation’.
“There’s nothing intuitive here,” Kollo said. “It’s pattern recognition.”
So over time, as the computer is fed more examples of numeral 8s, it converts these into vectors, long strings of numbers that to us mean nothing, but to the computer correlate to “8”. Eventually when it scans an image of an 8, no matter in whose handwriting, it will correctly identify it as such.
This kind of logic applies to all sorts of things. Computers use it to identify faces using 128 such vectors. That’s what a single person is to a machine: 128 probabilistic computations that to us look like very random batches of numbers. Faces are hard. Easier, then to turn credit scores, financial statements, earnings figures and financial forecasts into vectors and recreate them as probable matches – which can then be fed into an algorithm of buys and sells.
Restructuring buys and sells
In isolation, this may be happy news for quant investors, but it doesn’t change the business model. It might just be a better way of creating a trading call.
But Kollo says machine learning will upend how quant firms make market decisions.
“We will go from looking at financial statements or a company’s product, to a company’s tradeable network. What if a machine can recognize connections between entities that matter?”
In other words, quant investing – and by extension, many forms of active fund management – will no longer be about analyzing the company. It will be about how a company’s amassed vectors correlate to other companies’.
“A portfolio manager can track a hundred stocks,” Kollo said. “But not a hundred corporate relationships.”
But by turning company information – all of its numeral 8s – into machine-readable vectors, computers can then begin to recognize connections among companies that no human could spot. Not even managers of the companies themselves.
And as computers get faster, portfolio managers will create endless permutations of questions about a stock. That means far more ways to take original views on investment decisions, giving fund managers new ways to generate alpha.
Dream versus reality
That’s still science fiction. “Today we don’t know the rules well enough,” Kollo said of the technology. For now, more effort is going towards using machine learning to validate consensus economic theory and to understand how patterns are created, rather than trying to outsmart markets. The way that computers arrive at decisions, using logical transformation, is still not understood by programmers, and they want to get a grip on it before unleashing alpha-generating (?) monsters into the world.
“It’s a distrust of purely data-led investment,” Kollo said, noting again that Rosenberg isn’t a black-box firm.
That said, AXA Rosenberg is using its machine learning experience to introduce new products. Factor timing is first. Now that smart beta has been embraced by institutional clients, can the machine understand how long-term returns behave in specific time periods, and time these to avoid market losses? It’s a project that combines A.I. with the harvesting of massive amounts of data from all kinds of sources.
Many quants are working on this, and factor timing usually involves a variation of tactical asset allocation techniques, but Kollo believes that a trading strategy based on machine learning will outperform.
Another application is ESG funds, as environmental, social and governance mandates are big business now in Europe. From a quant’s perspective, machine learning can find correlations between ESG language in company statements and the degree to which these are implemented.
But does this information produce better alpha?
“I don’t know yet,” Kollo said.
Lu Global reverses the Lufax story
Lufax began as a P2P and became a wealth manager – in Singapore, it’s adding secondary trading.
Lu Global, a wealth-management fintech in Singapore, has just launched a marketplace to enable its customers to trade the same products they bought on the company’s website.
Kit Wong, CEO at Lu Global, says the company has developed its consumer-facing business and is now selling both funds and structured products.
But it believes some clients want to get out of these positions, particularly structured notes. Instead of having to hold them to maturity, they can now see if other users in the Lu system are willing to buy them (at a discount).
Wong says the firm, which has a capital markets services (CMS) license in Singapore, serves about 300,000 customers. Some are resident in Singapore (where the business can only market to accredited investors), others are from outside, who can be either professional investors or retail.
The biggest segment of investors are mainland Chinese, who already know the Lufax brand, but there are also a lot of Taiwanese and Hongkongers, and a growing number of Southeast Asian users, Wong says.
The electronic marketplace has just gone live, so it has no volumes to speak of. Lu Global does not take positions in this secondary trading environment – it merely matches its existing customer base in case users want to make trades among themselves.
Lu Global declined to state its assets under management. Wong says the largest number of products are mutual funds, issued by the likes of BlackRock and Pimco – but the biggest volumes are in structured products.
He believes this may have to do with economic and political uncertainty in the region, which is spurring demand for products with known outcomes and terms.
But such products only pay out upon maturity – and the same destabilizing factors may be leading more investors to want to cash out early, even if they do so at a loss. But providing a marketplace not only gives them access to liquidity (assuming there’s a buyer on the other side) but also lets them sell at a better rate.
The launch of this product is a strange parallel to parent Lufax’s journey. Shanghai-based Lufax began in 2011 as a peer-to-peer marketplace for transactions, financing, and investment management. It exited the transactions and financing aspects to focus just on wealth management.
Lu Global built itself first as a marketplace for wealth products – but now it’s expanding into secondary trading, creating a marketplace for customers to exchange financial products before they reach maturity among themselves – a different kind of P2P than lending, which mainland authorities are clamping down on.
Half of Invesco’s China sales now via digital
But as the PRC joint venture learns how to distribute digitally, Invesco remains unsure of robo’s role in Asia.
This week DigFin is highlighting three asset-management firms’ approach to digital distribution, particularly in China. See also strategies from AllianceBernstein and BEA Union Investments. Go here for more insights into digital asset and wealth management.
Invesco is using its joint venture in mainland China, Invesco Great Wall, to figure out digital distribution.
The business now manages about $50 billion of assets, of which about 80% is retail, making it the fourth-largest Sino-foreign fund house in China. Over the past two years, half of retail inflows have come from new digital channels, as opposed to the traditional reliance upon banks, says Andrew Lo, senior managing director and Asia-Pacific CEO of Invesco in Hong Kong.
This is in keeping with a broader trend in the global mutual funds industry, which is shifting from one based on products to one focused more on investment solutions. “There’s an emphasis on designing outcomes for clients, such as through asset allocation or structuring,” to combine types of risk and asset classes.
That’s driven both by client demand as well as market volatility and challenges to active fund houses to deliver alpha (outperformance) on a net-free basis, compared to ultra-affordable passive investments tracking a benchmark.
That’s been an emerging story for the funds industry over the past decade. But on top of that is a new wrinkle: the ability to use technology to speed up operations and to reach more people.
“Technology is now changing the distribution landscape,” Lo said. “In China, it’s having quite an impact on reaching retail investors.”
For now this has been a story unique to mainland China, where existing bank channels (which dominate funds distribution in most Asian markets) are not well developed, and where regulation favors digital disrupters like Ant Financial.
The power of digital was evident in Ant’s success with money-market funds (under an affiliated fund house, Tianhong Asset Management), but it has now extended to equity and quant products onshore – products that Invesco’s J.V. now sells through fintech channels, including Ant, East Money Information, JD.com (Jingdong) and Snowball Finance (Xueqiu).
This has not been straightforward, however. Fund management companies are designed to cater to bank distributors, and are built on old-fashioned tech.
“We learned how to do digital marketing,” Lo said. “It’s very different to traditional distribution. It’s iterative, it changes fast, and you have to listen to customer feedback.” Partnering with digital channels has also required a different sense of product design, and to rebuild the company’s operational process to support round-the-clock digital sales and support.
Lo says the experience will be increasingly relevant as other markets digitalize, although they may need to be tweaked, depending on local regulation, client behavior and distributor demands. “Some things we can learn and apply elsewhere as the world goes digital,” Lo said.
The onshore funds market manages about Rmb14 trillion (almost $2 trillion) in total assets among 135 asset managers authorized to sell to retail clients, of which are 44 Sino-foreign JVs.
But most of these JVs are run by the local partner, with foreign shareholders having less influence. They are limited to stakes no greater than 49%, and local partners are often banks or other powerful institutions. One analyst told DigFin that local fund houses are not particularly bold when it comes to digital channels; and even if they are, the lessons don’t flow to the foreign partner.
But Invesco Great Wall’s case is different. Both Invesco and Great Wall Securities own 49%, with two other shareholders holding another 1% each. Given that Great Wall Securities has its own in-house funds business, it has been willing to let Invesco drive the business. (Beijing has recently permitted J.P. Morgan Asset Management to take a 51% stake in its funds J.V.) Invesco Great Wall is also among the oldest funds JVs in China. It is today led by Shenzhen-based CEO Ken Kang Le.
In China, Invesco is leading the way in digital opportunities. Elsewhere it seems to be running with the rest of the herd. In the U.S. and the U.K., it has made digital acquisitions: Jemstep, a B2B robo-advisor that services U.S. bank distributors, and Inteliflo, a British platform to support financial advisors.
“We haven’t found the right use case in Asia,” Lo said. Onboarding a digital B2B (of B2B2C) platform needs scale, but Asia is fragmented, with each market requiring its own business and compliance needs.
“Digital transformation is still evolving,” Lo said. “My guess is it can be like it is in China, where it’s a real thing that has become a major part of the industry.” But what that looks like elsewhere remains hard to know – or at least hard for justifying a business case.
New China distribution not just for money-market funds
Investors on digital platforms are beginning to look to other products, says BEA Union’s Rex Lo.
This week DigFin is highlighting three asset-management firms’ approach to digital distribution, particularly in China. We will also provide strategies from Invesco and AllianceBernstein. Go here for more insights into digital asset and wealth management.
Retail investors in China accessing funds via digital platforms are beginning to diversify away from money-market funds. That is creating opportunities to push ETFs and active funds, says Rex Lo, managing director for business development at BEA Union Investments.
China’s retail funds industry is mainly about money-market funds (MMFs). The total industry size is Rmb13.2 trillion, or $1.9 trillion, of which MMFs account for 57%, or Rmb7.7 trillion.
Among MMFs, by far the biggest player is Tianhong Asset Management, whose product, Alibaba’s Yuebao fund, is Rmb1.2 trillion in size, or $162 billion – the largest money-market fund in the world.
It’s no surprise then that digital distribution platforms in China mainly cater to MMFs. Lo says until recently, MMFs accounted for about 80% of all funds sold on digital platforms. This is propelled businesses such as Tianhong (which of course is sold via Ant Financial) and a few bank-affiliated fund houses with big MMF products, such as CCB Principal and ICBC Credit Suisse.
But it has made digital distribution of limited interest for fund houses looking to sell equity funds or other actively managed products; for them, traditional distribution via banks has remained the only viable channel.
MMFs: less big
Lo thinks this is changing, however.
The popularity of MMFs lies mainly in the fact that they offered high returns combined with guarantees, real or assumed by investors – assumptions the government has been reluctant to upset.
Yuebao and other MMFs usually invest in non-standardized wealth-management products (themselves supposedly “guaranteed”, with investors assuming a government backstop), that returned 5% to 8% to those managers. They in turn offered investors 5%, an equity-like return on what’s meant to be an ultra-safe and liquid asset class.
Over the past few years, however, Chinese banking and securities regulators have been trying to shift the funds industry onto a footing that respects risk and return, and clamping down on the supply of shadow-banking instruments available to portfolio managers.
“Today MMFs return only a little over 2%, while A shares are doing well,” Lo said. “As demand for money market funds declines, turnover has fallen, so these distributors are now promoting index or active funds.”
In recent months, Lo says, MMFs account for only 70% of sales on digital channels, with ETFs now gaining ground.
Accessing the mainland market
BEA Union is able to sell its Hong Kong-domiciled Asia fixed-income fund to Chinese retail investors through a scheme called MRF, Mutual Recognition of Funds.
This program, which began in 2015, allows fund managers on either side of the border to sell eligible products through a master-agent arrangement. Regulators in mainland China have been slow to approve such funds, however, and there are only seven Hong Kong products available via MRF, including BEA Union’s (and 48 mainland funds available for sale in Hong Kong).
Lo is hoping to take advantage of the shifting fortunes among asset classes to use digital channels to push BEA Union’s bond fund.
Platforms such as Ant Financial are requesting the fund house for more material around equities and active funds management. It’s a big, long-term commitment to investor education – especially for foreign fund managers whose ranking is low on Ant Financial and other digital platforms.
“Domestic investors want familiarity,” Lo acknowledged. “But we continue marketing because we want to be on the platform. Today it’s more for exposure than real [inflows], and ticket sizes are as small as Rmb100 ($14). But if you have 100,000 investors, that becomes a lot of money.”
The intention of this ongoing marketing is to become sufficiently well known among Ant’s users to take advantage when retail investors want to invest overseas.
New ways of doing business
Adding platforms such as Ant to traditional distribution methods has been an eye-opener, Lo says. “They don’t think like a traditional finance company. They’re a fintech, so they’re very responsive and open to new ideas. And they’re independent – they’re not a bank with its own funds J.V. – so there aren’t conflicts of interest.”
Marketing was not the only part of business that had to adjust.
“I was amazed when we began to work with these firms,” Lo said. “Enhancements that would take months to get done in Hong Kong take them a few days. We can learn a lot from working with fintechs.” It’s knowledge that will come in handy as more banks in Hong Kong and Asia add mutual funds to their mobile trading apps, as Standard Chartered did earlier this year.
There are limits, however, to how far a fund house can go selling products on mainland China’s digital platforms.
Those channels are limited to funds from either local licensed retail-facing houses, or offshore products eligible via MRF. The retail funds market in China, at $1.9 trillion, is only a fraction of the total investments industry, which is about $9.7 trillion – but that includes separate licensed businesses for asset managers linked to insurance companies, or to trusts, or to banks. Those businesses for now can’t market to retail or use sell via e-commerce players.
BEA Union is a joint venture formed in 2007 between Bank of East Asia and Germany’s Union Investments. Its initial business model was to service local pension and insurance customers, so its investment expertise has been mainly in Asian fixed income. It has since developed funds in Hong Kong and Asia equities, and its total AUM is now $11.2 billion.
It is the only foreign fund house to establish a wholly owned foreign enterprise (Woofie) in Qianhai (part of Shenzhen), as opposed to Shanghai. This was partly because the authorities in Qianhai were very welcoming, and because Bank of East Asia has a presence in southern China, and the fund house hopes to take advantage of this should cross-border opportunities emerge (under the concept of a “Greater Bay Area”).
This medium-term ambition is another driver of BEA Union’s strategy to build an online brand on Ant Financial and other digital platforms.