Partisans of 'passive' or 'active' strategies could learn a thing or two tennis players, says Martin Steward.
Some people love nothing better than the trash-talking build-up to a heavyweight boxing bout. Being one of life's gentler souls, I much prefer the respectful sportsmanship that seems to have broken out over the last 10 years in men's tennis.
When the sportscaster asks Roger Federer how he is going to exploit Rafael Nadal's knee injury in the upcoming match, the answer is always a variation on: "Rafa is one of the best players in the world, and to challenge him you have to assume he is going to play well and focus on making sure that your own game is as good as it can be."
The people at Vanguard need to watch more tennis and less boxing. When I met with head of equities Sandip Bhagat recently, I had hoped to learn why he thought Vanguard's products were worth buying. Instead, he seemed intent on trash-talking the opposition - particularly fundamental indexation.
"Fundamental indexation is not as passive as it appears to be from its positioning and its labelling," he said. "It pursues its insight in a mechanical, automated manner - so it is a passive representation of something. But it is no longer the passive representation of the market; it's the passive representation of an active strategy.
"A less sophisticated investor may see that term, 'indexation', and assume that means they are getting the market, a 'smarter' market index - and therefore a free lunch in terms of better returns. That's something we want to provide some education and advocacy on."
I guess straw men are a hazard of trash-talking - but let's assume for the sake of argument there is a bunch of investors convinced that anything with the word 'index' in it is delivering 'passive' exposure to the market. What's the problem?
Bhagat argued that just as active strategies deliver risk above and beyond the basic risk of the equity market, so must passive representations of active strategies. The danger is that you think you are buying equity market risk, but in fact you are getting a tilt to a bunch of other risks. With fundamental indexation, Bhagat claimed that extra risk (which he defined as active risk against an equivalent market-cap index) is 75-80% value, 10-12% small-cap and 10-12% high-beta.
"Each of these themes are what I would call a risk premium, and investing in them requires the condition that you can bear the volatility in excess of what the equity market delivers," he said. "Value companies are those that are either in distress or chronically economically sensitive; they are either financials that make their money from the sloping yield curve or industrials like auto manufacturers with high operating leverage; making them vulnerable to pain in market down cycles […] Try absorbing an extra 10% of losses on top of the 30% losses from the equity market."
Yikes. Distress? Chronic economic sensitivity? An extra 10% of losses? This is far from a comprehensive scientific survey, but, according to Russell Investments' figures, while the Russell Global index (with dividends, priced in USD) fell 58.7% peak to trough between 31 October 2007 and 9 March 2009, the Russell Fundamental Global index (with dividends, priced in USD), constructed by Research Affiliates, fell 56.0%. One gets similar results for the Europe indices; the US indices are inseparable over that timeframe; and the emerging markets indices show cap-weighted outperforming by about 3 percentage points.
Still, I was happy to follow Bhagat quite a long way down this path of reasoning. I especially liked his point that one of the key aims of 'passive' investment products should be to deliver exposure to systematic risk premiums, and that they should therefore be priced low enough to reflect their mechanistic methodologies and pass as much of the premium as possible to the end investor. That's a great argument for making any 'index' product as cheap as possible.
Bhagat's main claim, however, was that fundamental indices are less 'passive' than cap-weighted indices, and should be sold as such. I was even prepared to go along with him on this, as long as we stuck with the proposition that cap-weighted indices represented the more 'passive' exposure to equity markets.
Whereas I was concerned to avoid the assumption that 'the market' was the same as 'the asset class', Bhagat seemed to use the terms interchangeably ("The market-cap index is the passive representation of the market or asset class opportunity"; "The equity market delivers a risk premium and the harvesting of that risk premium is a function of the investor's ability to bear the risk of that risky asset class"). That seemed telling to me, as it revealed the begged question at the heart of Bhagat's criticism of alternative weighting. A 'passive' exposure is an unbiased exposure to the equity asset class, and anyone who doesn't have that exposure is 'active'. But is the market-cap index 'passive' according to this definition?
The evidence can get very selective when these criticisms are thrown around. To make the argument that fundamental indexation was doing nothing but deliver systematic risk premiums, Bhagat himself observed that Fama and French had added the value and size factors to the CAPM framework decades ago. He chose not to recall that Carhart added momentum a few years later. If we accept that it is possible for a cap-weighted index to have biases against equity risk, momentum would be one of the big ones.
And this gets to the point about how fundamental indexation "positions" itself. It would be absurd to argue that Research Affiliates (say) claims that its fundamental indices have no risk-factor biases against cap-weighted indices (and by extension the market itself). Indeed, the point of them is that they do have such biases! The implication at the heart of most alternative-weighting methodologies is that biases against some theoretical 'true' equity risk can be minimised, relative to the huge biases against that theoretical risk embedded in the cap-weighted indices.
So what represents the 'true' equity risk? Bhagat's criticism of fundamental indexation's tilt to value and small-caps indicates that he accepts two things: (1) Equity risk is an aggregate of individual (diversifying) risks; and (2) 'Passivity' is defined by lack of bias against those risks (rather than, say, buying-and-holding with zero or low turnover). But accepting those premises one can convincingly argue that the passive equity risk exposure is the equal (unbiased) exposure to each of the individual risks available from the equity market, rather the equity market itself (which has big biases toward certain of those individual risks). That would presumably be the tangency portfolio in the CAPM framework - which decades of research has shown not to be the cap-weighted portfolio.
At the very least, we should concede it is contentious to point to the biases in competitors' portfolios while ignoring those in your own.
This finally brings us back to Bhagat's points about "positioning" and "labelling". If I had thought I was buying 'equity market risk' with my index product in 1999, I could have had absolutely no complaint when TMT became 40% of my holdings a year or so later. If that's what the market is buying, that's what the market is buying. If I'd thought I was buying 'equity risk', however, I'd have been severely miffed.
I suggest that Bhagat confused these categories because, rather than focusing on "educating' investors and "advocating" the cap-weighted index on its own terms, he was too focused on trash-talking the opposition.
That is a great shame because Vanguard, specializing in equilibrium, buy-and-hold, cap-weighted index portfolios, can minimise friction costs and offer savers super-cheap exposure to equity market risk. And that is a genuine social good, worthy of advocacy.