Month: February 2016

The Strength of Strong-Form

The Efficient-market hypothesis, developed by Eugene Fama, is probably one of the most contested topics in the history of financial economics. Both on an empiric and theoretical basis economists, traders and media pundits often claim to have the truth that either supports or undermines the observations of Fama dating back to the 1960s. The EMH states that asset market prices reflect all available information and hence obey a ‘random walk’ on a ‘fat-tailed’ distribution.

But when financial economists say ‘reflect all available information’ what exactly do they mean? There are 3 possible interpretations:

Weak form- traded assets reflect all past public information.
Semi-strong form- traded assets reflect all past information but also adjust instantaneously to reflect all present public information.
Strong form- traded assets, in addition to having the properties of semi-strong form also reflect current and past ‘insider information’.
To most people, if one of the above forms is the most plausible it is either the weak or the semi-strong. By my intuition at least, it seems implausible that markets are able to completely price information that cannot even be acted on by most agents. Eugene Fama himself describes access to information on a continuum, where at first the cost is quite flat before quickly skewing off to infinity as more information is obtained on the margin. Hence, by his own admissions it may seem that ‘insider information’ is too costly to attain for most agents.

Nevertheless, Fama and others in favour of the EMH insist that more statistical evidence was found in favour of the strong strand of EMH than for the weaker forms. How can this be? I am by no means well-trained enough to devour the reams of statistical evidence on both sides of the debate and hence will refrain from offering a general opinion on whether the EMH is inherently flawed or a beautiful prophecy; but I will attempt to explain how the Strong-Form of the EMH has more going for it than is commonly assumed. In fact, it seems that Strong-Form is to be expected from some classic financial economics theorems which expose the vulnerabilities of using ‘economic intuitions’ alone. Contrary to popular belief, in this vein a “strong-form” of the EMH is probably more consistent with highly volatile markets than the weaker versions.

The clue to understanding why the ‘strong-form’ may be empirically sound comes from an unlikely source. Nothing really needs to be said about investors obtaining fundamental information for themselves. Rather, it has something to do with the effect that asset trading itself has on the dispersion of information.

A formal and somewhat abstract way of thinking about price information dispersal throughout a market comes from the No-trade theorem. Informally, this theorem states that under conditions whereby the market is composed completely of rational traders (no-noise) and operating at an efficient equilibrium there is no scope to profit off of private information. This is because given that all traders are rational with acknowledgement of collective rationality, any attempt to initiate a trade will reveal the private agent’s informational advantage and hence change market expectations in line with that. Consequently, all insider information is automatically reflected in prices as if it were public information.

Note that the assumptions behind the No-trade theorem are obviously not met by reality, and that theorems themselves are only strictly true when their assumptions are met. However, when thinking about information diffusion in an economy notice the game theoretic nature of asset trading in the hyper-rational world of the No-trade theorem.

Implicitly there is some strategic interaction, where each agent attempts to anticipate the motives of the others, and due to strict rationality ends up concluding that their own information is less accurate. If any forms of the EMH are to be taken seriously then there is the inevitable recognition roughly speaking the market will be at some kind of ‘perceived’ efficient equilibrium with noise traders not having the financial capital to issue the largest trades.

Hence, whilst the assumptions of the No-trade theorem are not completely met even when the EMH holds, the No-trade theorem still provides a context for understanding how Strong-Form may actually be more plausible than expected.

The idea is that using Fama’s model of an information-cost continuum, only agents with access to substantial capital have the capacity or incentive to actually obtain that information and then trade on it. Following this, the trade using insider information is likely to be quite large, often issued by a known Hedge Fund which can ‘lever’ the information up to the hilt. The fact that the private information holder is only likely to exist if they are agents with an ability to trade large volumes implies that they are often sophisticated agents whose valuations are more likely to be perceived as fundamental rather than noise by the rest of the market. Thus, using the game theoretic analysis from the No-trade theorem it would be quite rational for other agents to reprice their own valuations in line with the offer/bid by the private information holder.

It can be expected that in a world of fast paced traders trailing behemoth funds there is little to be gained from trading on insider information which serves as a signal more than anything else. Practically, it may even be in the public’s interest to permit insider trading where there is scope to maintain the information-cost continuum (via an excise tax perhaps); seeing as it can signal an efficient pricing mechanism.

Furthermore, the existence of these ‘game theoretic’ patterns whereby smaller traders attempt to piggy-back off the information of more sophisticated ones in a world of Strong-Form EMH could also be more prone to volatility of the sort first identified by Mandelbrot. In this context, the ‘fat tails’ of financial markets can represent the cascading of less sophisticated agents attempting to bid immediately after large sophisticated ‘shock trades’. In turn, quick readjustments may be the realisations that high-frequency ‘trailing’ is often misplaced since large trades may simply reflect internal dynamics of the leading firm. This fits the observable facts that shocks are often random, and readjustments follow seemingly unimportant information.

Note, that none of the above discussion is bullet-proof financial theory, nor is it an exposition of the strength of the EMH. Rather, I have attempted to explain practically how insider information can be priced effectively in markets, and how it is linked to seemingly unexplainable volatility that may otherwise be used as evidence for irrationality.