Efficient markets or not

Students have recently been asking about how anyone can continue to defend the efficient markets hypothesis (EMH) against the backdrop of the financial crisis and volatility we’ve seen in the last couple of years.

I think there are two useful ways of looking at this. One is suggested by the famous behavioural economist Richard Thaler, in an article in the Financial Times on August 5, 2009. He summarised the EMH as having two messages: i) no free lunches; and ii) the price is right. The first of these is broadly intact; it is still prudent to assume that you are unlikely to beat the market without a great deal of skill and hard work. But ii) is in some jeopardy, as it is hard to believe that the prices of securities in the last couple of years were always reflecting their true fair value.

Secondly it’s worth revisiting the late Paul Samuelson’s 1998  idea that stock markets are “micro efficient” but “macro inefficient”. Samuelson was one of the greatest of 20th century economists and one who spent enough time in the stock market to be a little sceptical about some of the more extreme EMH ideas. One reason for why the EMH might not hold so well for the whole market is that it cannot be arbitraged in the way that individual stocks can. Here a relevant example is the technology, media and telecoms bubble of the late 1990s (often referred to as the dot-com bubble, but the biggest damage was done in the telecoms sector). Many practitioners correctly saw that we were in a bubble, where valuations were increasingly absurd and money was moving in a speculative way. But nobody knew when it would end and some previously successful hedge fund managers lost a lot of money in being right – but too early.

Interesting Thaler says in his FT article that with all of their flaws, financial markets remain the best way to allocate capital. But we should not assume the price is right.

6 Responses to Efficient markets or not

  1. I think the interesting question here is how we actually arrive at price in a given market. Take gold for example. It seems to me that since 2007, the price of the asset reflects market concern over monetary easing. People fear that the aggressive expansion of the Fed’s balance sheet will result in hyperinflation not seen since the Weimar Republic. Personally, I can understand why that fear would be “priced into” the value of gold, but I think it reflects a misunderstanding of monetary policy during a liquidity trap more than does fundamentals. Gold prices may be “correct” in that it’s signaling concern. If hyperinflation does not occur, then prices contain an “incorrect” signal. Regardless, if the money that is being poured into gold is coming from the liquidation of assets xyz, are the resulting prices of xyz “correct”?

    Hypothetically, if investors hold only two assets – gold or xyz – then as investors substitute towards gold, the price of xyz should fall. If normally xyz’s price reflects true value, then the current price should not: xyz = True Value – Flight to Gold. If xyz happen to be bank stocks, then if the new “price” is too low, the bank will not have enough capital & will subsequently fail. Interestingly enough, this phenomena seemed to occur with Latin American sovereign debt after Lehman failed. In a flight to quality, most investors redirected their money towards US sovereign debt. The impact was so strong, that Latin American countries considered issuing USD denominated debt because the demand for their local denominated assets flew from otherwise unchanged economies. (One can make the argument of LatAm growth being levered to US growth, with new yields simply reflecting this relationship – I believe the truth is in the middle)

    So, how should we think about prices? What is “true value”? To be rigorous here: an asset’s price represents the perception of value derived from the liquidity surrounding that asset. If you’re holding a bond to maturity, then the price you charge is likely the probability of default plus some benchmark riskless rate. Since you’re not actively trading, the change in yield of similar instruments should have little bearing on the value of your asset. The only determinant of price is your perception, expressed here as interest demanded for risk. Ironically enough, accounting may useful for economic theory (Held Till Maturity vs. Mark to Market). However, the minute you trade, the perception of the individuals you trade with – in particular the marginal buyer – make all the difference. When there’s little liquidity or visible free cash flow to temper one’s biases, then the value of perception increases. The perception of decoupling may very well explain why oil prices skyrocketed in the summer of 2007.

    This first was made clear to me when I was on sell-side portfolio strategy, talking to a REITs client. The client claimed that the underlying fundamentals of his space were strong, but that a couple of hedge funds were messing it up. Traditionally, his space was small & the buyers present understood the assets. Suddenly, a new brand of buyer entered the market with one strategy – buy up CDS, short REITs. It was as if a bunch of hooligans broke into an art gallery and started arguing that a Picasso is worth less than their child’s finger-painting.

    We hope this inefficiency is removed after repeated iterations, when knowledgeable investors recognize and exploit the misprice (mean reversion). However, it may not happen – especially if there are barriers to entry. You may be the voice of reason, but the ignorant may have a megaphone (or in this case, synthetic forwards and a more liberal mandate than that of institutional investors). If traders are Bayesian updaters, then rumors – however untrue – may become stubborn, stylized facts that alter the behavior of an asset.

    Ultimately, as you said, we hope that free markets efficiently allocate capital. Here, I’m sympathetic to Hayek’s “The Use of Knowledge in Society”. “True Value” in the way we hope should allocate capital perfectly. However, distortions in the structure of markets strongly impact price discovery. I feel we’ve focused less on how markets are organized and more on how a given idea of markets optimize. To quote, “This character of the fundamental problem has, I am afraid, been obscured rather than illuminated by many of the recent refinements of economic theory, particularly by many of the uses made of mathematics … It seems to me that many of the current disputes with regard to both economic theory and economic policy have their common origin in a misconception about the nature of the economic problem of society.” (Hayek, 1945)

    Thoughts?

  2. Excellent post!

    I completely agree with the point that financial markets are “micro efficient” but “macro inefficient”. As an arbitrage trader, I continually look at any mis-pricing opportunities among financial instruments, but it’s impossible to arbitrage against the mis-pricing of the whole market.

  3. I have two questions-

    1)Is the market not the sum total of all individual stocks (which have been arbitraged)? And if all individual stocks reflect rational prices on account of being arbitraged, why should the phenomenon differ for the market as a whole?

    2)When the market is “macro-inefficient”, wont the inefficiency, in a sense, trickle down, distort individual stock prices and make these individual stocks “micro-inefficient”?

  4. Reading this reminds me of my late room mate. That guy was one of the smartest characters I know, but he was a little beatnik for my tastes though. Anyways I delighted in reading this, thanks. Will give me something to discuss when I see him.

  5. EMH is clearly not all bad. while it is certainly not the most effective model in the light of recent financial crisis there are elements to it which are applicable and realistic.

Leave a reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

We are using cookies on our website

Are you happy to accept our analytics cookies, which help us learn about our website visitors and their use of this site? Learn how to disable all cookies.