Tuesday 1 December 2009

The Famous Efficient Market Hypothesis II:

Chronicle of a Death Foretold.

I went these days through some old paper clippings. You know: you read something, find it interesting, and decide to keep it. But you never use it again, and even forget about it.

Somehow that stuff never gets thrown away. And, surprisingly, one day it turns out it was a good idea to keep it, after all.

Re-reading those old clippings and photocopies, I had a clear feeling of déjà vu. Check this out:

"It is difficult to pick up a newspaper these days without seeing another article about a major company that has taken an unexpected financial loss due to derivatives transactions gone awry".

"That's not too old. Probably a 2007-2008 newspaper article", you might say. "Surely it talks about investment banks affected by sub-prime mortgages?"

Nope. This is over 14 years old: it's from a 1995 issue of the Harvard Business Review [0]. They are not talking about investment banks affected by sub-prime mortgages. The corporate casualties were in the manufacturing and mining/oil industries: Procter & Gamble Co, Metallgesellschaft AG, and Kashima Oil.

Some big financial players also had their fingers burnt around those years: Askin Capital (a group of hedge funds ominously specialising in mortgages), PaineWebber (mutual funds management) and Kidder, Peabody & Co (an investment bank). And, as the article was published, Barings Bank (which would go belly up later in 1995) had just started its way to doom.

Derivatives, for those unfamiliar, are financial "products" such as options, general swaps, CDS, CDO and CMO. They are called derivatives because their value "derives" from the value of an asset, subject to chance variations.

I will spare you of the statistical details, but to give you an idea of the complexity involved: until recently economists were not considered capable to understand derivatives.

So, the financial industry hired physicists and mathematicians (so-called rocket scientists, derivative geeks, or simply quants,): "The migration of physicists to the financial industry that started as a trickle in the late 1970s has now [1994] become a steady stream (…) In short, these days Wall Street, when compared to physics [departments], is Club Med".[1].

And quants were paid handsomely, too: apart from a very generous base pay, they were given bonuses, according to the profitability of the derivatives they designed and sold: "Last year [1993] times were good: derivatives experts fretted about how to avoid paying too much tax on bumper bonuses" [2]. And profitability depended to a large extent on the derivatives' complexity, which was just fine, with quants and their employers.

According to the same article, by 1994, though, things had changed for the worse for quants: "The travails of Procter & Gamble, a consumer goods manufacturer, and other American companies that lost money on esoteric swaps have deterred investors from buying anything unusual"

In fact, partly preceding these companies' "travails", partly reinforced by them, those years witnessed much soul-searching from the academic community and the financial industry.

To be sure, not all that soul-searching was radical: some was focused on the way derivatives should be used, as opposed to how they were used by these unfortunate companies. In a later article we'll discuss this in more detail.

Here we'll deal with the more radical criticism, directed not only against derivatives, but in general against financial theory as a whole.

A Little Bit of History.

What follows draws on an extraordinarily useful and informative 1993 article by Nancy A. Nichols [3], which I cannot possibly recommend enough (I apologize, though, for the detour).

Nichols' account dispels some misunderstandings related to how widely accepted was the Efficient Market Hypothesis (EMH) in financial practice (see my previous article in this Famous Financial Market Hypothesis series).

Nichols provides a succinct account of the historical development of the three main pillars of modern financial theory:

(1) Portfolio Theory by Prof. Harry Markowitz (early 1950s): the notion that an investor who does not diversify can't do better than an investor who does. As the market is already diversified, investors on average can't do better than the market.

(2) EMH by Prof. Eugene Fama (early 1960s). This hypothesis is equivalent to the Rational Expectations Hypothesis, developed around the same time by the late Prof. John Fraser Muth: "outcomes do not differ systematically (i.e. regularly or predictably) from what people expected them to be" [4]. Because errors are not systematic (i.e. they are random and averaging 0), they cannot be predicted: i.e. they are immediately factored in the prices of assets. In other words: there are no such things as bubbles.

(3) CAPM by Prof. William Sharpe and others (early to mid 1960s): used "to define the unique risk of holding stocks in general and then to judge the risk of any one stock, in relation to the market as a whole", assuming that the EMH was valid.

However, from the mid 1980s, empirical work (ironically, some of it by Prof Fama himself) demonstrated systematic and persistent anomalies incompatible with the EMH and CAPM.

For a while, these anomalies were ignored as mere curiosities. Eventually, though, the empirical evidence forced a reconsideration of the theory:

"Yet today [1993] (…) that belief [that all business is quantifiable and that markets can be studied scientifically] is under attack from all sides: from those who say finance uses the wrong scientific paradigm to those who say finance isn't a science at all but an art".

Nichols identified three main objector groups:

(1) The Revisionists: individuals like Prof. Robert J. Shiller, who "don't advocate throwing out the theories that make up modern finance, but they sure would like to tinker with them".

(2) The "Gadflies": individuals like the late Prof. Louis Lowenstein [5] and Samuel M. Miller, who "eschew the scientific approach altogether, arguing that investors aren't always rational and that managers' constant focus on the markets is ruining corporate America".

(3) The Chaos Cabal: individuals like J. Doyne Farmer and Norman Packard, who proposed the use of heterodox tools, coming from physics and complexity theory, to the financial markets.

However by 1996 - barely three years after Nichols' article - The Economist's editors were not only satisfied that the Chaos Cabal's efforts were entirely futile, but in an interesting example of non sequitur (ignoring both the Revisionists, the "Gadflies" or the entirely unacknowledged Post Keynesians) concluded that the EMH had prevailed: "Many have now concluded that formal chaos theory has nothing practical to offer (…) And such objections will come as no surprise to financial economists who believe that the idea of predicting price movements contradicts the theory that markets are more or less efficient". [6] My emphasis.

So, in one simple stroke all criticism directed towards the EMH and CAPM, in the first instance (but also against modern finance theory, by extension) was dismissed and we could happily go on with our lives.

Back to Derivatives.

After that long detour, you might be asking yourself what does it have to do with derivatives?

Check this quote (yes, yet another one): "[Derivatives] trading volume is well over $10 trillion - approaching the combined gross national product of the US, Japan and Europe - most of it unregulated because government agencies have yet to catch up. Increasingly, observers have begun to worry that a major misstep could vaporize financial markets." [7]

But modern financial theory was okay, wasn't it? The EMH and to a lesser extent the CAPM had prevailed at the end. That was The Economist's institutional opinion, shared by many in the financial industry. No wonder, for example, Prof. Bhagwan Chowdhry assuaged SciAm's concerns thus:

"Probably not. For one, the amount of money at risk is usually only a tiny fraction of the trading volume - as little as a few thousand dollars on a $100-million deal. For another, unlike real markets, derivative markets are zero-sum: for every big loser, there is also a big winner. Unless a player defaults (with debts exceeding assets), wealth can only be redistributed, not created or destroyed". My underlined.

That's the magic of old papers: a tragically wrong opinion that took into account, nonetheless, some of the main elements that 13 years later would indeed cause a financial meltdown of worldwide proportions.


[0] Harvard Business Review. Perspectives section. January-February 1995. Page 33.

[1] Scientific American. Science and Business section: Wall Street. October 1994. Page 126.

[2] The Economist. Derivatives section: Pain and Gain. 09/07/1994. Page 80.

[3] Nancy A. Nichols. "Efficient? Chaotic? What's the New Finance?" Harvard Business Review. March-April 1993. Page 50.

[4] Thomas J. Sargent. "Rational Expectations". The Concise Encyclopedia of Economics.

[5] Dennis Hevesi. "Louis Lowenstein, Professor of Business Law and Critic of Wall Street, Dies at 83". The New York Times, 25/04/2009.

[6] The Economist. Finance and Economics section: Chaos under a Cloud. 13/01/1996. Page 69.
[7] Scientific American. The Analytical Economist. Derivatives: Not the Real Thing. January 1995. Page 28.

No comments:

Post a Comment