At the end of 2008, the government of Mexico hedged its crude oil exports. They hedged not at the market top of $145 but well into the collapse at $70. They then hedged their 2010 exports at $55, near the bottom of the market.
In 2007, money managers were chasing higher returns with leverage and high beta stocks. Then after a 50% collapse in equity prices they reduced their equity exposure and bought puts to hedge.
Barrick Gold Corp recently announced that they would be issuing new equity to buy back their gold hedges because they believe gold prices will continue rising. This comes after one of the most bullish decades for gold in history (gold was $270 at the start of 2001).
What do these examples have in common? Money managers, corporate titans, and industry experts are always chasing the market. They become more bullish after long rallies and hedge after significant sell offs. This is a tautology if you appreciate that the cause of a long rally is a growing and gradual appreciation of bullish fundamentals. By the time the major players are all confident that prices will continue rising, everyone who can buy has already bought so there are no new buyers to send prices higher. The peak of a rally is then by definition a peak in the confidence that prices will go higher.
Unlike in any other field, a consensus of experts can therefore be an inherent contrarian indicator. In medicine, every additional doctor I find who agrees with the diagnosis lends me additional confidence. In investing, every new analyst who believes prices must go higher instead weakens the case, because that is one more person who is already long.
Is this true of all investment consensus? A consensus that prices will remain stable is itself stable. Such a belief requires no new converts to be proven true. But a consensus that prices will rise must eventually be disappointed because rising prices require new money for buying; eventually we run out of new buyers or new money for buying.
A potential exception is equities where a population may continously devote a portion of newly created wealth to purchases. For example, if we all take 10% of our paychecks to buy equities every month, there is a continual influx of new money. As long as the general wealth of the society is growing, new money will outpace the withdrawals of retirees. The problem we faced in 2007 is that the commitment to devote savings to equity purchases was based on the belief that equities would rise. As equities fell and wealth was destroyed, workers reduced their exposure. In commodities, the situation is even starker. A consensus that prices will rise increases production putting further pressure on prices. Then when the bubble bursts, not only is demand reduced, but production has been enlarged so prices fall below the initial equilibrium.