Sunday, September 08, 2013

MSCI Quality Index

I was unaware MSCI had beaten AQR to the punch by producing a boatload of quality indices last spring.  These are applied worldwide, so they are necessarily more parsimonious than AQRs...but jeez, these are really barebones:

1) Net Income/Book Equity
2) Debt/Book Equity
3) Earnings volatility over 5 years

Instructively, they Winsorize the data, which everyone should do to financial ratios (ie, truncate extremums).  But, book equity in the denominator?  Earnings volatility over 5 years? Those seem like bad choices, and AQR's quality index will be superior.

I have a feeling MSCI is a bit confused, as they have another tab noting their 'Risk Premia Indexing', which they note
An accumulating body of empirical research has found positive gross excess returns from exposure to factors (or risk premia) such as Value, Momentum, Low Size (small firms), and Low Volatility stocks. The studies show that these factors historically have improved return-to-risk ratios. Today, interest in risk premia (also known as smart beta or alternative beta) has been widespread across the institutional investor community.
In other words, risk premia are really return premiums, because predictable returns only come from risk (in theory). But then, they also 'improve return-to-risk ratios', because, as we all know, these factors aren't risk in any obvious way, so strangely they all have 'excess return' premia. Indeed, 'value and size were initially thought to be due to distress risk, which would show up only episodically.  Alas, 'quality' is basically an metric of anti-distress, and this generates a return premium, which MSCI occasionally calls a 'risk premium' basically whatever asset outperforms over the next 20 year period will ex post be declared risky.  

The risk-begets-return model of economics is clearly nonfalsifiable amongst current financial academics and their coterie.  They still say interesting things on occasion, so it doesn't render them useless, but it definitely impairs their ability to see and interpret reality.


Anonymous said...

Each time “the market” is sliced and diced it 1) reduces the size of the business opportunity to manage other people’s money and 2) suggests a desire to address a growing perception that a product is not living up to the dreams of investors. Vanguard started out offering access to “the stock market”, DFA reduced the size of that business opportunity by making portfolio construction conditional on value. AQR and others then defined an even smaller conditional product space by throwing in momentum. If DFA and AQR rationally add “profitability” to their already existing processes then the size of their business product spaces should roughly decline in half.

Mercury said...

I think the other takeaway is simply that running an index is rapidly becoming one of the best business models in the asset management business. R&D can’t be that expensive which at this point sounds a lot like back-testing and cherry-picking metrics then putting them in an attractive package. If you come up with one that gains any kind of traction it can be a real cash cow. Larger firms often pay two and three times per index, per user for the right to use the same data on different platforms.

Also, if you didn’t see it, Olympic wrestling news on CNBC:

James said...

Your old friend nassim is now in now accusing a good friend of plagiarism, even though he cited his work 23 times

Anonymous said...

Why are those choices poor?