Factor funds ‘do right things for wrong reasons’ – Intech

Firm says conventional investing wisdom is missing out on alpha

Intech, a $50 billion quant manager, says factor investors have got it wrong.

The firm has long trodden a different path to other systematic investors, building strategies based solely on the effects of diversification, even as the idea that a handful of stock characteristics – ‘factors’ – drive returns has become ubiquitous among quants. Now Intech is challenging that consensus head on.

The firm says popular factor-based strategies work, not because they target recognised factors such as value, but because – unintentionally – through creating more diversified portfolios, they tap a deeper source of alpha.

“Factor funds are doing the right thing but for the wrong reasons,” David Schofield, president of Intech’s international division, tells Risk.net.

If Intech is right, factor strategies can be improved; they could also stop working.

Diversity by accident

“All of the size effect, and most of the value effect and low-volatility anomaly” – core tenets of factor theory – are attributable to selecting and periodically rebalancing a diversified portfolio, Schofield says.

“Factor portfolios are picking up this diversification effect inadvertently, accidentally and certainly not in a targeted, efficient, risk-controlled way. Once you acknowledge this true alpha source, which is common to the various factor approaches, it makes sense to optimise your portfolio to that specific effect, and to try to eliminate the residual risk factors that are just a source of tracking error and can lead to long periods of underperformance.”

To make the point, in a recent paper, Intech authors alongside Johannes Ruf from the London School of Economics, repeated an exercise from a 2013 study that sought to show why wholly random strategies could beat the market.

Using factors, the earlier study failed to explain all of the “surprising” alpha in the strategies. But Intech was able to tie the outperformance fully to the higher level of diversification in the random portfolios compared with a cap-weighted index. 

David Schofield

Factor experts ascribe the outperformance of factor portfolios to outperformance in the individual stocks they contain.

Instead, the source of outperformance is the impact of diversification on long-term returns, with portfolios that are less volatile relative to the stocks they hold generating higher compound returns, all else being equal, Intech argues.

“A portfolio’s compound return is the compound return of all the stocks plus an extra bit. That extra bit – in mathematical terms the portfolio’s ‘excess growth rate’ – is the boost to return that comes from diversification,” Schofield says. “And that extra bit can be meaningful.”

The excess growth rate depends only on correlations and volatility, yet it explains much of the above-market return of many systematic strategies, he says.

In the paper, the authors calculated the average compound returns of the top 1,000 US stocks over 48 years by their market cap ranking, and found the annualised averages for smaller stocks were essentially the same as for bigger stocks – results that refute the existence of a size premium as set out in factor theory.

‘Perpetually amazed’

Intech has built its strategies around the effects of diversification since it was founded by Princeton mathematician Robert Fernholz in 1987.

Its strategies are based on rebalancing portfolios to maintain a target level of diversification, which generates a trading profit that accounts for most of the relative outperformance.

Close to 90% of the firm’s strategies have beaten their benchmark over the five years to June, though the past three years have seen weaker performance, with US markets in particular being driven by a small group of high-growth stocks. 

Intech’s ideas are far from mainstream in traditional finance circles, but are central to mathematical finance, Schofield says. He describes himself as “perpetually amazed” at the “mystery” of traditional investors still being unaware of many of the ideas the firm draws on. 

But neither is Intech a voice in the wilderness.

Tobam – an asset manager launched by Yves Choueifaty, the former chief executive of Credit Lyonnais Asset Management –  also follows a strategy of optimising diversification.

Like Intech, Choueifaty sees much of the value of factor strategies as resulting from being more diversified than their benchmark indexes. And the growing popularity of such strategies risks making them less diversified, he says.

“Factor investors would say these stocks reward risk better than others. I would say these stocks have rewarded risk better than others, in the past, which may continue. But saying one risk factor will outperform – that it will be cheap – forever is difficult to understand. The only thing that will outperform forever is being well diversified because the index – by loading on the more popular stocks – is expensive,” Choueifaty says.

Factor fans not convinced

Other quants speaking to Risk.net say it would take more to convince them to abandon their reliance on factors.

Some question whether diversification alone can explain the performance of factor strategies beyond size and value, such as momentum for example.

They acknowledge, however, that size is one of the weaker factors so far established, saying it becomes more pronounced for stocks outside the top 1,000, perhaps explaining Intech’s research results. 

And even those sceptical about ditching their factor models think diversification strategies might soon have their day.

Seen through a conventional factor lens, these strategies favour value and small-sized stocks, says Vitali Kalesnik, head of equity research at Research Affiliates and one of the authors of the 2013 paper.

“Value stocks are trading cheap versus growth stocks. The spreads between value and growth are as wide in some markets as they have ever been. As likely as not there’s going to be a value bull market for the next five years,” he says.

At the same time, some diversification strategies are much less capacity constrained compared with funds targeting specific factor exposures, he adds.

Research Affiliates itself runs a set of fundamental indexes that aim to redress the inherent bias of cap-weighted benchmarks and are also seen as a play on diversification.

Back at Intech, Schofield emphasises that factor portfolios can outperform for long periods but underperform for equally long periods, pointing to the slump in value strategies in some recent years as an example.

Diversification is the “underlying alpha source” of many factor strategies, he argues, but – perhaps unlike some factors – the diversification effect is “unlikely to be arbitraged away”.