Reinventing Asset Allocation: Investment Solutions
How Portfolio X and a Devil’s Advocate Can Solve Complex Problems in an Uncertain World
It’s late 2017. For most of the year, financial markets have been very strong. There are some signs that the bull market is nearing its end — including ongoing uncertainty around the fate of the president’s tax cut legislation — but even still, few investors actually pull back as the year draws to a close.
What kept investors from pulling back a few months earlier? Certainly, optimism was a big factor. But as important as any market dynamic was another reason: a tendency on the part of investors to interpret new evidence as confirmation of existing beliefs or theories.
Sure enough, in February 2018 markets plummeted. The reasons weren't terribly surprising: inflation fears in the wake of a better-than-expected jobs report, political uncertainty around government spending. And in the days following, things bounced back. But the spike in volatility was frightening: At one point, the Dow industrials fell more than 1,500 points in one day — the biggest single-day drop in market history.1
Behavioral scientists call it “confirmation bias,” and it’s a topic that Dan Farley, the head of State Street Global Advisors’ Investment Solutions Group, thinks about a lot. It’s not surprising. In a world where political decisions and the media are increasingly driven by social media and the “fake news” phenomenon, even the simplest decisions can be fraught with peril. Complicated financial decisions like where and when to invest are even more difficult.
Dan’s team has a big job. They’re entrusted with managing US$307 billion in client assets, with investors looking for a range of services — from a top-down perspective on their investment portfolios, to portfolio construction at the widest lens, to narrower mandates to boost returns on one slice of their overall portfolio.
What is one of the biases his team runs into most? “The need to keep hiring additional active managers,” Dan says. While active management is a critical aspect of reaching return objectives, he says, many institutions believe that the more managers they hire, the more diversified their portfolio will be. But according to Dan, the evidence suggests otherwise. “Obviously, diversification is really important to reducing risk. And it is not that active management doesn’t play an important role in meeting objectives. But at a certain point, simply putting more of your money into competing active equity management philosophies doesn’t increase returns. They begin to offset each other. The result is that it just becomes brown soup — an expensive brown soup.”
Here, Mehvish Ayub, a London-based senior investment manager, steps in. “Tactical asset allocation is a way of solving this bias toward adding more security selection managers.” By focusing on asset allocation, she says, the team can exploit some of the macro inefficiencies in the market. This allows them to offer clients a tool that not only generates alpha, but also alpha that is distinct (or, in investing parlance, “uncorrelated”) from stock picking and other traditional types of active management.
Building the Perfect Beast: Portfolio X
But what’s the process for getting there? Each month, the team refines what it calls “Portfolio X.” Far from the images of superheroes its name conjures, Portfolio X is the Investment Solutions Group’s master portfolio. Mehvish puts it another way: The portfolio represents the team’s “best ideas and point of view.”
Every step of the team’s monthly alpha process has a quantitative and fundamental component — a way of working through biases of their own. The meticulous process begins by assessing the market environment, combining a proprietary quantitative-based market regime indicator with a qualitative assessment of the political environment provided by the team’s Politics and Policy group. “The view on Chinese steel tariffs is going to look very different in Sydney than here in London,” says Mehvish. “This gives us a sense of the risk-seeking appetite investors have.”.
Next, the model looks at market return drivers and produces expected returns for every asset class in 140 markets. Each return is checked and rechecked by asset class teams comprised of portfolio managers located around the globe. “Good ideas travel,” Mehvish says. “By recognizing that early in the process, we try to identify what the model might be missing.”
The model proposes a series of optimized trades that can be used in a portfolio, then the group breaks into seven independent teams that make trade recommendations. The structure of multiple voting teams with a firewall between each of them is designed to avoid herd mentality and groupthink. “Instead of listening to the loudest person in the room,” she says, “we’re relying on the wisdom of crowds.”
Enter the Devil’s Advocate
But there is one more step to take. It’s something called the Devil’s Advocate — and according to Mehvish, it is a unique aspect of the process.
“The Devil’s Advocate was actually one of the things I found attractive about this job when I interviewed,” the London School of Economics graduate says. It’s not hard to understand why: At a point in the process when most shops would be taking the product to market, State Street Global Advisors challenges members of its team to make the best possible arguments against the consensus trades the team has voted to make.
Mehvish, who has served as a Devil’s Advocate a number of times during her nearly four years with the firm, describes the rapid-fire process. Kicking the tires on some of the team’s biggest assumptions, she says the Devil’s Advocate goes back to the beginning of the alpha process. “We’ll analyze the quant side — build arguments on why we shouldn’t have trusted the decision on that model. We’ll pressure-test the qualitative decisions the teams have made. Or maybe it’s a macro factor — views for or against the dollar.”
The role is challenging. “By design, you aren’t told the exact topic until the day before,” she says. “You really need to think on your feet and ask, ‘What could substantiate this argument?’ ‘What thematics does the base portfolio represent?’ ‘What could be challenged?’ You have to review the work of the asset class teams and third-party research to ensure the rigor of the arguments.”
Mehvish notes that, “Often we do our own evaluation of quant models based on the idea that maybe the team is missing something.” Are there ever any hard feelings? “No comment,” she says with a smile, though she quickly adds that the position of Devil’s Advocate constantly rotates. “At some point, we all will have been in this position, which I think helps build trust in the process.”
How successful is the Devil’s Advocate? Mehvish estimates that these arguments and insights are incorporated into the final product about 20 to 25 percent of the time. If that seems low, it’s a testament to the quality of the process it is challenging. “The entire alpha process is structured in a considered way,” she says. “If the process works and the trades are sound to begin with, then they should stand up, whatever arguments are brought against them.”
Driving Results for Clients
After all of that work, there’s scarcely a dollar of the US$100 billion the tactical asset allocation team manages that isn’t influenced in some way by Portfolio X.
Dan describes a foundation that recently turned to State Street Global Advisors for some help. “They were looking for total returns of 7.5 percent. But even a reasonably prudent strategic asset allocation will only get you into the high fives based on our long-term expected returns. By the time we were asked to help, they were already spending 4.5 percent of their endowment each year, which needs to grow by 2 percent to combat inflation. They spent another 1 percent on fees — and already had a broad lineup of active managers.”
Can adding an additional active manager help make up the difference? “To an extent,” Dan says. “But most institutional investors can only dedicate so much of their portfolio to active equities or a single investment manager. This is where tactical asset allocation can give them another option.”
It’s been an effective tool for investors, according to Dan. “The baseline for our model portfolio is 80 bps of return above benchmark,” he says. “We’ve done a pretty good job delivering that on an annualized basis over the last decade.” Looking back, Dan and Mehvish agree that the whole process is ultimately about playing Devil’s Advocate and challenging biases. “It forces us to pause and think about different counterpoints and weigh them,” Mehvish says.
At the same time, Dan adds, it’s a good reminder that there are limits to how much objectivity managers can add to the investment process. “You can’t eliminate bias any more than you can eliminate risk,” he says. But, he quickly notes, “one of the things we’ve learned is that by really understanding those biases and how they impact the decisions people make, you can turn them into an opportunity.”